FLUX.2 Klein 9B FP8 (Recommended) vs FLUX.2 Klein 4B (Apache 2.0)
Last data update: 2026-06-11T22:17:24.345Z
Intent: image-same-family-cost-and-size-comparison
Quality gate
- Eligible for indexable canary
- yes
- Evidence score
- 7
| Check | Pass | Evidence |
|---|---|---|
| same-type intent | yes | image vs image |
| distinct models | yes | flux2-klein-9b-fp8 vs flux2-klein-4b |
| pricing present | yes | $0.39/hr and $0.39/hr |
| deployment target present | yes | gpu-l4 and gpu-l4 |
| source config facts present | yes | 39 facts and 39 facts |
| same-type catalog context present | yes | 27 image catalog rows |
| asset evidence when workflow model | yes | 64 matched asset rows |
Evidence summary
- flux2-klein-9b-fp8__flux2-klein-4b intent: image-same-family-cost-and-size-comparison; this pair is compared as a same-type deployment decision, not as an arbitrary generated adjacency;
- flux2-klein-9b-fp8__flux2-klein-4b cost profile: flux2-klein-9b-fp8 costs $0.39/hr and $280.8/mo at 720h; flux2-klein-4b costs $0.39/hr and $280.8/mo at 720h;
- flux2-klein-9b-fp8__flux2-klein-4b instance profile: flux2-klein-9b-fp8=gpu-l4 medium comfyui-flux2; flux2-klein-4b=gpu-l4 small comfyui-flux2;
- flux2-klein-9b-fp8__flux2-klein-4b description comparison: flux2-klein-9b-fp8=Fastest Flux model; Sub-second generation, 4-step distilled; Runs on RTX 4090;; flux2-klein-4b=Fully open source (Apache 2.0); Fast generation on RTX 3090/4070; ~13GB VRAM;
- flux2-klein-9b-fp8__flux2-klein-4b group evidence: flux2-klein-9b-fp8=FLUX.2 Klein is the fastest model in the Flux family; The 9B FP8 variant delivers sub-second generation in 4 steps, while the 4B model is fully open source under Apache 2.0 and runs on consumer GPUs;; flux2-klein-4b=FLUX.2 Klein is the fastest model in the Flux family; The 9B FP8 variant delivers sub-second generation in 4 steps, while the 4B model is fully open source under Apache 2.0 and runs on consumer GPUs;
- flux2-klein-9b-fp8__flux2-klein-4b nearby catalog count: 27 image models were available for internal context during page selection;
- flux2-klein-9b-fp8__flux2-klein-4b matched deployment assets: flux-2-klein-9b-fp8.safetensors, qwen_3_8b_fp8mixed.safetensors, flux2_dev_fp8mixed.safetensors, mistral_3_small_flux2_fp8.safetensors, flux1-dev-fp8.safetensors, flux1-dev-kontext_fp8_scaled.safetensors, flux1-kontext-dev-fp8.safetensors, flux2-dev-Q4_K_M.gguf;
| FLUX.2 Klein 9B FP8 (Recommended) | FLUX.2 Klein 4B (Apache 2.0) | |
|---|---|---|
| Model group | flux2-klein | flux2-klein |
| Type | image | image |
| Source id | flux2-klein-9b-fp8 | flux2-klein-4b |
| Model size tier | medium | small |
| Deployment template | comfyui-flux2 | comfyui-flux2 |
| Docker image base | comfyui | comfyui |
| Container port | 8188 | 8188 |
| Web UI port | n/a | n/a |
| Volume mount path | /data | /data |
| Parameters | 9B | 4B |
| Context window | n/a | n/a |
| VRAM (GB) | 24 | 24 |
| Recommended GPU | gpu-l4 | gpu-l4 |
| GPU type | NVIDIA L4 | NVIDIA L4 |
| GPU count | 1 | 1 |
| vCPU | 12 | 12 |
| RAM (GB) | 50 | 50 |
| Disk (GB) | 200 | 200 |
| Instance type | gpu-efficient | gpu-efficient |
| RunPod id | n/a | n/a |
| $/hr | $0.39 | $0.39 |
| $/month (24/7) | $280.8 | $280.8 |
| Popularity (1-10) | 11 | 10 |
| Deploy time (min) | 18 | 18 |
- Cost delta /hr
- $0
- VRAM delta GB
- 0
- Best GPU tier to fit both
- gpu-l4
Cost scenarios
| Usage level | Hours / month | flux2-klein-9b-fp8 | flux2-klein-4b | Delta |
|---|---|---|---|---|
| Prototype, 40 hr/mo | 40 | $15.6 | $15.6 | $0 |
| Part-time app, 160 hr/mo | 160 | $62.4 | $62.4 | $0 |
| Always-on, 720 hr/mo | 720 | $280.8 | $280.8 | $0 |
Decision matrix
| Factor | flux2-klein-9b-fp8 | flux2-klein-4b | Winner |
|---|---|---|---|
| Lower hourly cost | $0.39/hr | $0.39/hr | flux2-klein-9b-fp8 |
| Lower 24/7 monthly cost | $280.8/mo | $280.8/mo | flux2-klein-9b-fp8 |
| Lower VRAM tier | 24GB | 24GB | flux2-klein-9b-fp8 |
| Higher popularity score | 11/10 | 10/10 | flux2-klein-9b-fp8 |
| Faster deployment estimate | 18 min | 18 min | flux2-klein-9b-fp8 |
| Template specificity | comfyui-flux2 | comfyui-flux2 | flux2-klein-9b-fp8 |
Pair config delta
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__slug; flux2-klein-9b-fp8=flux2-klein-9b-fp8; flux2-klein-4b=flux2-klein-4b; comparison=different value; source=adapter identity
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__sourceId; flux2-klein-9b-fp8=flux2-klein-9b-fp8; flux2-klein-4b=flux2-klein-4b; comparison=different value; source=modelConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__displayName; flux2-klein-9b-fp8=FLUX.2 Klein; flux2-klein-4b=FLUX.2 Klein; comparison=same value; source=modelConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__sizeVariant; flux2-klein-9b-fp8=9B FP8 (Recommended); flux2-klein-4b=4B (Apache 2.0); comparison=different value; source=modelConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__modelGroup; flux2-klein-9b-fp8=flux2-klein; flux2-klein-4b=flux2-klein; comparison=same value; source=modelConfig.js / modelSeoData.js
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__type; flux2-klein-9b-fp8=image; flux2-klein-4b=image; comparison=same value; source=modelConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__modelSize; flux2-klein-9b-fp8=medium; flux2-klein-4b=small; comparison=different value; source=resolved model config
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__templateName; flux2-klein-9b-fp8=comfyui-flux2; flux2-klein-4b=comfyui-flux2; comparison=same value; source=resolved model config
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__dockerImageBaseName; flux2-klein-9b-fp8=comfyui; flux2-klein-4b=comfyui; comparison=same value; source=resolved model config
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__containerPort; flux2-klein-9b-fp8=8188; flux2-klein-4b=8188; comparison=same numeric value; source=resolved deployment config
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__webUIPort; flux2-klein-9b-fp8=not-specified; flux2-klein-4b=not-specified; comparison=same value; source=resolved deployment config
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__volumeMountPath; flux2-klein-9b-fp8=/data; flux2-klein-4b=/data; comparison=same value; source=resolved deployment config
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__recommendedInstance; flux2-klein-9b-fp8=gpu-l4; flux2-klein-4b=gpu-l4; comparison=same value; source=instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__recommendedInstanceName; flux2-klein-9b-fp8=GPU Efficient (L4); flux2-klein-4b=GPU Efficient (L4); comparison=same value; source=instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__recommendedInstanceSpecs; flux2-klein-9b-fp8=7B-13B LLMs, SDXL Images, Power Efficient; flux2-klein-4b=7B-13B LLMs, SDXL Images, Power Efficient; comparison=same value; source=instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__gpuType; flux2-klein-9b-fp8=NVIDIA L4; flux2-klein-4b=NVIDIA L4; comparison=same value; source=instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__gpuCount; flux2-klein-9b-fp8=1; flux2-klein-4b=1; comparison=same numeric value; source=instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__vramGB; flux2-klein-9b-fp8=24; flux2-klein-4b=24; comparison=same numeric value; source=instanceConfig.js derived fit
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__vCPU; flux2-klein-9b-fp8=12; flux2-klein-4b=12; comparison=same numeric value; source=instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__ramGB; flux2-klein-9b-fp8=50; flux2-klein-4b=50; comparison=same numeric value; source=instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__diskGB; flux2-klein-9b-fp8=200; flux2-klein-4b=200; comparison=same numeric value; source=instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__instanceType; flux2-klein-9b-fp8=gpu-efficient; flux2-klein-4b=gpu-efficient; comparison=same value; source=instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__runpodInstanceId; flux2-klein-9b-fp8=not-specified; flux2-klein-4b=not-specified; comparison=same value; source=instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__hourlyCost; flux2-klein-9b-fp8=0.39; flux2-klein-4b=0.39; comparison=same numeric value; source=instanceConfig.js deployCost
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__monthlyCost720h; flux2-klein-9b-fp8=280.8; flux2-klein-4b=280.8; comparison=same numeric value; source=derived from hourly cost
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__deployTimeMinutes; flux2-klein-9b-fp8=18; flux2-klein-4b=18; comparison=same numeric value; source=adapter estimate
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__popularity; flux2-klein-9b-fp8=11; flux2-klein-4b=10; comparison=modelA higher by 1; source=modelConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__groupPopularity; flux2-klein-9b-fp8=99; flux2-klein-4b=99; comparison=same numeric value; source=modelConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__paramCount; flux2-klein-9b-fp8=9B; flux2-klein-4b=4B; comparison=different value; source=parsed from model metadata
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__contextWindow; flux2-klein-9b-fp8=not-specified; flux2-klein-4b=not-specified; comparison=same value; source=parsed from model metadata
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__description; flux2-klein-9b-fp8=Fastest Flux model; Sub-second generation, 4-step distilled; Runs on RTX 4090;; flux2-klein-4b=Fully open source (Apache 2.0); Fast generation on RTX 3090/4070; ~13GB VRAM;; comparison=different value; source=modelConfig.js / modelSeoData.js
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__useCases; flux2-klein-9b-fp8=Sub-second image generation; Real-time AI applications; Consumer GPU deployment; Open-source commercial use (Apache 2.0); flux2-klein-4b=Sub-second image generation; Real-time AI applications; Consumer GPU deployment; Open-source commercial use (Apache 2.0); comparison=same value; source=modelSeoData.js
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__relatedGroups; flux2-klein-9b-fp8=flux-2; z-image-turbo; flux; flux2-klein-4b=flux-2; z-image-turbo; flux; comparison=same value; source=modelSeoData.js
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__sourceFiles; flux2-klein-9b-fp8=src/config/modelConfig.js; src/config/modelSeoData.js; src/config/instanceConfig.js; flux2-klein-4b=src/config/modelConfig.js; src/config/modelSeoData.js; src/config/instanceConfig.js; comparison=same value; source=snapshot sync
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__aboutBlurb; flux2-klein-9b-fp8=FLUX.2 Klein is the fastest model in the Flux family; The 9B FP8 variant delivers sub-second generation in 4 steps, while the 4B model is fully open source under Apache 2.0 and runs on consumer GPUs;; flux2-klein-4b=FLUX.2 Klein is the fastest model in the Flux family; The 9B FP8 variant delivers sub-second generation in 4 steps, while the 4B model is fully open source under Apache 2.0 and runs on consumer GPUs;; comparison=same value; source=model config or SEO snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__configKeys; flux2-klein-9b-fp8=flux2-klein-9b-fp8, flux2-klein-4b, flux2-klein-9b; flux2-klein-4b=flux2-klein-9b-fp8, flux2-klein-4b, flux2-klein-9b; comparison=same value; source=model config or SEO snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__defaultConfig.containerPort; flux2-klein-9b-fp8=8188; flux2-klein-4b=8188; comparison=same numeric value; source=resolved config snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__defaultConfig.dockerImageBaseName; flux2-klein-9b-fp8=comfyui-flux; flux2-klein-4b=comfyui-flux; comparison=same value; source=resolved config snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__defaultConfig.modelSize; flux2-klein-9b-fp8=medium; flux2-klein-4b=medium; comparison=same value; source=resolved config snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__defaultConfig.templateName; flux2-klein-9b-fp8=comfyui-flux; flux2-klein-4b=comfyui-flux; comparison=same value; source=resolved config snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__defaultConfig.volumeMountPath; flux2-klein-9b-fp8=/data; flux2-klein-4b=/data; comparison=same value; source=resolved config snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__displayName; flux2-klein-9b-fp8=FLUX.2 Klein; flux2-klein-4b=FLUX.2 Klein; comparison=same value; source=model config or SEO snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__dockerImageBaseName; flux2-klein-9b-fp8=comfyui; flux2-klein-4b=comfyui; comparison=same value; source=model config or SEO snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__dockerImageTag; flux2-klein-9b-fp8=flux2; flux2-klein-4b=flux2; comparison=same value; source=model config or SEO snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__groupPopularity; flux2-klein-9b-fp8=99; flux2-klein-4b=99; comparison=same numeric value; source=model config or SEO snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__h1; flux2-klein-9b-fp8=Deploy FLUX.2 Klein; flux2-klein-4b=Deploy FLUX.2 Klein; comparison=same value; source=model config or SEO snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__id; flux2-klein-9b-fp8=flux2-klein-9b-fp8; flux2-klein-4b=flux2-klein-4b; comparison=different value; source=model config or SEO snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__modelGroup; flux2-klein-9b-fp8=flux2-klein; flux2-klein-4b=flux2-klein; comparison=same value; source=model config or SEO snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__modelSize; flux2-klein-9b-fp8=medium; flux2-klein-4b=small; comparison=different value; source=model config or SEO snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__modelType; flux2-klein-9b-fp8=image; flux2-klein-4b=image; comparison=same value; source=model config or SEO snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__name; flux2-klein-9b-fp8=FLUX.2 Klein 9B FP8; flux2-klein-4b=FLUX.2 Klein 4B; comparison=different value; source=model config or SEO snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__popularity; flux2-klein-9b-fp8=11; flux2-klein-4b=10; comparison=modelA higher by 1; source=model config or SEO snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__recommendedInstance.deployCost; flux2-klein-9b-fp8=0.39; flux2-klein-4b=0.39; comparison=same numeric value; source=resolved config snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__recommendedInstance.disk; flux2-klein-9b-fp8=200; flux2-klein-4b=200; comparison=same numeric value; source=resolved config snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__recommendedInstance.gpuCount; flux2-klein-9b-fp8=1; flux2-klein-4b=1; comparison=same numeric value; source=resolved config snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__recommendedInstance.gpuType; flux2-klein-9b-fp8=NVIDIA L4; flux2-klein-4b=NVIDIA L4; comparison=same value; source=resolved config snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__recommendedInstance.id; flux2-klein-9b-fp8=gpu-l4; flux2-klein-4b=gpu-l4; comparison=same value; source=resolved config snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__recommendedInstance.name; flux2-klein-9b-fp8=GPU Efficient (L4); flux2-klein-4b=GPU Efficient (L4); comparison=same value; source=resolved config snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__recommendedInstance.ram; flux2-klein-9b-fp8=50; flux2-klein-4b=50; comparison=same numeric value; source=resolved config snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__recommendedInstance.suitableFor; flux2-klein-9b-fp8=7B-13B LLMs, SDXL Images, Power Efficient; flux2-klein-4b=7B-13B LLMs, SDXL Images, Power Efficient; comparison=same value; source=resolved config snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__recommendedInstance.type; flux2-klein-9b-fp8=gpu-efficient; flux2-klein-4b=gpu-efficient; comparison=same value; source=resolved config snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__recommendedInstance.vCPU; flux2-klein-9b-fp8=12; flux2-klein-4b=12; comparison=same numeric value; source=resolved config snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__relatedGroups; flux2-klein-9b-fp8=flux-2, z-image-turbo, flux; flux2-klein-4b=flux-2, z-image-turbo, flux; comparison=same value; source=model config or SEO snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__resolvedConfig.containerPort; flux2-klein-9b-fp8=8188; flux2-klein-4b=8188; comparison=same numeric value; source=resolved config snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__resolvedConfig.dockerImageBaseName; flux2-klein-9b-fp8=comfyui; flux2-klein-4b=comfyui; comparison=same value; source=resolved config snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__resolvedConfig.modelSize; flux2-klein-9b-fp8=medium; flux2-klein-4b=small; comparison=different value; source=resolved config snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__resolvedConfig.templateName; flux2-klein-9b-fp8=comfyui-flux2; flux2-klein-4b=comfyui-flux2; comparison=same value; source=resolved config snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__resolvedConfig.volumeMountPath; flux2-klein-9b-fp8=/data; flux2-klein-4b=/data; comparison=same value; source=resolved config snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__sizeVariant; flux2-klein-9b-fp8=9B FP8 (Recommended); flux2-klein-4b=4B (Apache 2.0); comparison=different value; source=model config or SEO snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__slug; flux2-klein-9b-fp8=flux-2-klein; flux2-klein-4b=flux-2-klein; comparison=same value; source=model config or SEO snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__templateName; flux2-klein-9b-fp8=comfyui-flux2; flux2-klein-4b=comfyui-flux2; comparison=same value; source=model config or SEO snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__useCases; flux2-klein-9b-fp8=Sub-second image generation, Real-time AI applications, Consumer GPU deployment, Open-source commercial use (Apache 2.0); flux2-klein-4b=Sub-second image generation, Real-time AI applications, Consumer GPU deployment, Open-source commercial use (Apache 2.0); comparison=same value; source=model config or SEO snapshot
- flux2-klein-9b-fp8__flux2-klein-4b field=flux2-klein-9b-fp8__flux2-klein-4b__fact__workflowUrl; flux2-klein-9b-fp8=https://docs.comfy.org/tutorials/flux/flux-2-klein; flux2-klein-4b=https://docs.comfy.org/tutorials/flux/flux-2-klein; comparison=same value; source=model config or SEO snapshot
Description comparison
| Field | flux2-klein-9b-fp8 | flux2-klein-4b |
|---|---|---|
| Description | Fastest Flux model; Sub-second generation, 4-step distilled; Runs on RTX 4090; | Fully open source (Apache 2.0); Fast generation on RTX 3090/4070; ~13GB VRAM; |
| Use cases | Sub-second image generation; Real-time AI applications; Consumer GPU deployment; Open-source commercial use (Apache 2.0) | Sub-second image generation; Real-time AI applications; Consumer GPU deployment; Open-source commercial use (Apache 2.0) |
| Related groups | flux-2; z-image-turbo; flux | flux-2; z-image-turbo; flux |
Deployment facts
- flux2-klein-9b-fp8__flux2-klein-4b source id comparison: flux2-klein-9b-fp8=flux2-klein-9b-fp8; flux2-klein-4b=flux2-klein-4b;
- flux2-klein-9b-fp8__flux2-klein-4b template comparison: flux2-klein-9b-fp8 uses comfyui-flux2 with comfyui; flux2-klein-4b uses comfyui-flux2 with comfyui;
- flux2-klein-9b-fp8__flux2-klein-4b instance comparison: flux2-klein-9b-fp8 runs on GPU Efficient (L4) at $0.39/hr and 24GB VRAM; flux2-klein-4b runs on GPU Efficient (L4) at $0.39/hr and 24GB VRAM;
- flux2-klein-9b-fp8__flux2-klein-4b fit notes: flux2-klein-9b-fp8=7B-13B LLMs, SDXL Images, Power Efficient; flux2-klein-4b=7B-13B LLMs, SDXL Images, Power Efficient;
- flux2-klein-9b-fp8__flux2-klein-4b use case comparison: flux2-klein-9b-fp8=Sub-second image generation; Real-time AI applications; Consumer GPU deployment; Open-source commercial use (Apache 2.0); flux2-klein-4b=Sub-second image generation; Real-time AI applications; Consumer GPU deployment; Open-source commercial use (Apache 2.0);
- flux2-klein-9b-fp8__flux2-klein-4b source files: flux2-klein-9b-fp8=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js; flux2-klein-4b=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js;
Deployment payload fields
flux2-klein-9b-fp8__flux2-klein-4b.candidate_a.modelType=image flux2-klein-9b-fp8__flux2-klein-4b.candidate_a.modelName=flux2-klein-9b-fp8 flux2-klein-9b-fp8__flux2-klein-4b.candidate_a.instanceId=gpu-l4 flux2-klein-9b-fp8__flux2-klein-4b.candidate_a.templateName=comfyui-flux2 flux2-klein-9b-fp8__flux2-klein-4b.candidate_a.containerPort=default flux2-klein-9b-fp8__flux2-klein-4b.candidate_a.dockerImageBaseName=comfyui flux2-klein-9b-fp8__flux2-klein-4b.candidate_b.modelType=image flux2-klein-9b-fp8__flux2-klein-4b.candidate_b.modelName=flux2-klein-4b flux2-klein-9b-fp8__flux2-klein-4b.candidate_b.instanceId=gpu-l4 flux2-klein-9b-fp8__flux2-klein-4b.candidate_b.templateName=comfyui-flux2 flux2-klein-9b-fp8__flux2-klein-4b.candidate_b.containerPort=default flux2-klein-9b-fp8__flux2-klein-4b.candidate_b.dockerImageBaseName=comfyui
Matched model assets
- flux2-klein-9b-fp8__flux2-klein-4b asset-1 filename=flux-2-klein-9b-fp8.safetensors; size=9.43GB; type=unet; destination=diffusion_models; repository=black-forest-labs/FLUX.2-klein-9b-fp8; host=huggingface.co; url=https://huggingface.co/black-forest-labs/FLUX.2-klein-9b-fp8/resolve/main/flux-2-klein-9b-fp8.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-2 filename=qwen_3_8b_fp8mixed.safetensors; size=4.4GB; type=clip; destination=text_encoders; repository=Comfy-Org/flux2-klein; host=huggingface.co; url=https://huggingface.co/Comfy-Org/flux2-klein/resolve/main/split_files/text_encoders/qwen_3_8b_fp8mixed.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-3 filename=flux2_dev_fp8mixed.safetensors; size=32.2GB; type=unet; destination=diffusion_models; repository=Comfy-Org/flux2-dev; host=huggingface.co; url=https://huggingface.co/Comfy-Org/flux2-dev/resolve/main/split_files/diffusion_models/flux2_dev_fp8mixed.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-4 filename=mistral_3_small_flux2_fp8.safetensors; size=18GB; type=clip; destination=text_encoders; repository=Comfy-Org/flux2-dev; host=huggingface.co; url=https://huggingface.co/Comfy-Org/flux2-dev/resolve/main/split_files/text_encoders/mistral_3_small_flux2_fp8.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-5 filename=flux1-dev-fp8.safetensors; size=11.9GB; type=unet; destination=unet; repository=Kijai/flux-fp8; host=huggingface.co; url=https://huggingface.co/Kijai/flux-fp8/resolve/main/flux1-dev-fp8.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-6 filename=flux1-dev-kontext_fp8_scaled.safetensors; size=11.09GB; type=unet; destination=diffusion_models; repository=Comfy-Org/flux1-kontext-dev_ComfyUI; host=huggingface.co; url=https://huggingface.co/Comfy-Org/flux1-kontext-dev_ComfyUI/resolve/main/split_files/diffusion_models/flux1-dev-kontext_fp8_scaled.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-7 filename=flux1-kontext-dev-fp8.safetensors; size=11.9GB; type=unet; destination=unet; repository=6chan/flux1-kontext-dev-fp8; host=huggingface.co; url=https://huggingface.co/6chan/flux1-kontext-dev-fp8/resolve/main/flux1-kontext-dev-fp8-e4m3fn.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-8 filename=flux2-dev-Q4_K_M.gguf; size=20.1GB; type=unet_gguf; destination=unet; repository=city96/FLUX.2-dev-gguf; host=huggingface.co; url=https://huggingface.co/city96/FLUX.2-dev-gguf/resolve/main/flux2-dev-Q4_K_M.gguf
- flux2-klein-9b-fp8__flux2-klein-4b asset-9 filename=flux2-dev-Q5_K_M.gguf; size=24.1GB; type=unet_gguf; destination=unet; repository=city96/FLUX.2-dev-gguf; host=huggingface.co; url=https://huggingface.co/city96/FLUX.2-dev-gguf/resolve/main/flux2-dev-Q5_K_M.gguf
- flux2-klein-9b-fp8__flux2-klein-4b asset-10 filename=flux2-dev-Q8_0.gguf; size=35GB; type=unet_gguf; destination=unet; repository=city96/FLUX.2-dev-gguf; host=huggingface.co; url=https://huggingface.co/city96/FLUX.2-dev-gguf/resolve/main/flux2-dev-Q8_0.gguf
- flux2-klein-9b-fp8__flux2-klein-4b asset-11 filename=flux2-dev.safetensors; size=64.4GB; type=unet; destination=diffusion_models; repository=black-forest-labs/FLUX.2-dev; host=huggingface.co; url=https://huggingface.co/black-forest-labs/FLUX.2-dev/resolve/main/flux2-dev.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-12 filename=flux2-vae.safetensors; size=336MB; type=vae; destination=vae; repository=Comfy-Org/flux2-dev; host=huggingface.co; url=https://huggingface.co/Comfy-Org/flux2-dev/resolve/main/split_files/vae/flux2-vae.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-13 filename=mistral_3_small_flux2_bf16.safetensors; size=36GB; type=clip; destination=text_encoders; repository=Comfy-Org/flux2-dev; host=huggingface.co; url=https://huggingface.co/Comfy-Org/flux2-dev/resolve/main/split_files/text_encoders/mistral_3_small_flux2_bf16.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-14 filename=t5xxl_fp8_e4m3fn_scaled.safetensors; size=5.16GB; type=clip; destination=clip; repository=comfyanonymous/flux_text_encoders; host=huggingface.co; url=https://huggingface.co/comfyanonymous/flux_text_encoders/resolve/main/t5xxl_fp8_e4m3fn_scaled.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-15 filename=t5xxl_fp8_e4m3fn.safetensors; size=4.89GB; type=clip; destination=clip; repository=comfyanonymous/flux_text_encoders; host=huggingface.co; url=https://huggingface.co/comfyanonymous/flux_text_encoders/resolve/main/t5xxl_fp8_e4m3fn.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-16 filename=ae.safetensors; size=335MB; type=vae; destination=vae; repository=black-forest-labs/FLUX.1-dev; host=huggingface.co; url=https://huggingface.co/black-forest-labs/FLUX.1-dev/resolve/main/ae.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-17 filename=clip_l.safetensors; size=246MB; type=clip; destination=clip; repository=comfyanonymous/flux_text_encoders; host=huggingface.co; url=https://huggingface.co/comfyanonymous/flux_text_encoders/resolve/main/clip_l.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-18 filename=flux1-dev.safetensors; size=23.8GB; type=unet; destination=unet; repository=black-forest-labs/FLUX.1-dev; host=huggingface.co; url=https://huggingface.co/black-forest-labs/FLUX.1-dev/resolve/main/flux1-dev.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-19 filename=flux1-fill-dev.safetensors; size=23.8GB; type=unet; destination=unet; repository=black-forest-labs/FLUX.1-Fill-dev; host=huggingface.co; url=https://huggingface.co/black-forest-labs/FLUX.1-Fill-dev/resolve/main/flux1-fill-dev.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-20 filename=flux1-kontext-dev.safetensors; size=23.8GB; type=unet; destination=unet; repository=black-forest-labs/FLUX.1-Kontext-dev; host=huggingface.co; url=https://huggingface.co/black-forest-labs/FLUX.1-Kontext-dev/resolve/main/flux1-kontext-dev.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-21 filename=flux1-redux-dev.safetensors; size=182MB; type=style; destination=style_models; repository=black-forest-labs/FLUX.1-Redux-dev; host=huggingface.co; url=https://huggingface.co/black-forest-labs/FLUX.1-Redux-dev/resolve/main/flux1-redux-dev.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-22 filename=flux1-schnell.safetensors; size=23.8GB; type=unet; destination=unet; repository=black-forest-labs/FLUX.1-schnell; host=huggingface.co; url=https://huggingface.co/black-forest-labs/FLUX.1-schnell/resolve/main/flux1-schnell.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-23 filename=gemma_3_12b_it_fp8_e4m3fn.safetensors; size=12.8GB; type=clip; destination=text_encoders; repository=Lightricks/LTX-Video-2-0.9.5-DEV; host=huggingface.co; url=https://huggingface.co/Lightricks/LTX-Video-2-0.9.5-DEV/resolve/main/gemma_3_12B_it_fp8_e4m3fn.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-24 filename=hidream_i1_dev_fp8.safetensors; size=17.1GB; type=unet; destination=diffusion_models; repository=Comfy-Org/HiDream-I1_ComfyUI; host=huggingface.co; url=https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/resolve/main/split_files/diffusion_models/hidream_i1_dev_fp8.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-25 filename=hidream_i1_fast_fp8.safetensors; size=17.1GB; type=unet; destination=diffusion_models; repository=Comfy-Org/HiDream-I1_ComfyUI; host=huggingface.co; url=https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/resolve/main/split_files/diffusion_models/hidream_i1_fast_fp8.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-26 filename=hidream_i1_full_fp8.safetensors; size=17.1GB; type=unet; destination=diffusion_models; repository=Comfy-Org/HiDream-I1_ComfyUI; host=huggingface.co; url=https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/resolve/main/split_files/diffusion_models/hidream_i1_full_fp8.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-27 filename=ip-adapter_flux.safetensors; size=1.58GB; type=ip_adapter; destination=ipadapter; repository=InstantX/FLUX.1-dev-IP-Adapter; host=huggingface.co; url=https://huggingface.co/InstantX/FLUX.1-dev-IP-Adapter/resolve/main/ip-adapter.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-28 filename=llama_3.1_8b_instruct_fp8_scaled.safetensors; size=9.08GB; type=clip; destination=clip; repository=Comfy-Org/HiDream-I1_ComfyUI; host=huggingface.co; url=https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/resolve/main/split_files/text_encoders/llama_3.1_8b_instruct_fp8_scaled.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-29 filename=llava_llama3_fp8_scaled.safetensors; size=4.9GB; type=clip; destination=text_encoders; repository=Comfy-Org/HunyuanVideo_repackaged; host=huggingface.co; url=https://huggingface.co/Comfy-Org/HunyuanVideo_repackaged/resolve/main/split_files/text_encoders/llava_llama3_fp8_scaled.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-30 filename=ltx-2-19b-dev-fp8.safetensors; size=9.59GB; type=checkpoint; destination=checkpoints; repository=Lightricks/LTX-Video-2-0.9.5-DEV; host=huggingface.co; url=https://huggingface.co/Lightricks/LTX-Video-2-0.9.5-DEV/resolve/main/ltx-video-2-0.9.5-dev-fp8.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-31 filename=ltx-2-19b-distilled-fp8.safetensors; size=27.1GB; type=checkpoint; destination=checkpoints; repository=Lightricks/LTX-2; host=huggingface.co; url=https://huggingface.co/Lightricks/LTX-2/resolve/main/ltx-2-19b-distilled-fp8.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-32 filename=ltx-2.3-22b-dev-fp8.safetensors; size=29.2GB; type=checkpoint; destination=checkpoints; repository=Lightricks/LTX-2.3-fp8; host=huggingface.co; url=https://huggingface.co/Lightricks/LTX-2.3-fp8/resolve/main/ltx-2.3-22b-dev-fp8.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-33 filename=ltx-2.3-22b-distilled-fp8.safetensors; size=29.5GB; type=checkpoint; destination=checkpoints; repository=Lightricks/LTX-2.3-fp8; host=huggingface.co; url=https://huggingface.co/Lightricks/LTX-2.3-fp8/resolve/main/ltx-2.3-22b-distilled-fp8.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-34 filename=qwen_2.5_vl_7b_fp8_scaled.safetensors; size=9.38GB; type=clip; destination=text_encoders; repository=Comfy-Org/Qwen-Image_ComfyUI; host=huggingface.co; url=https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/text_encoders/qwen_2.5_vl_7b_fp8_scaled.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-35 filename=qwen_3_4b-engineer-v4-fp8.safetensors; size=4.02GB; type=clip; destination=text_encoders; repository=SeeSee21/Z-Anime; host=huggingface.co; url=https://huggingface.co/SeeSee21/Z-Anime/resolve/main/text_encoder/qwen_3_4b-engineer-v4-fp8.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-36 filename=qwen_image_edit_2509_fp8_e4m3fn.safetensors; size=19.03GB; type=unet; destination=diffusion_models; repository=Comfy-Org/Qwen-Image-Edit_ComfyUI; host=huggingface.co; url=https://huggingface.co/Comfy-Org/Qwen-Image-Edit_ComfyUI/resolve/main/split_files/diffusion_models/qwen_image_edit_2509_fp8_e4m3fn.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-37 filename=qwen_image_edit_2509_fp8mixed.safetensors; size=19.12GB; type=unet; destination=diffusion_models; repository=Comfy-Org/Qwen-Image-Edit_ComfyUI; host=huggingface.co; url=https://huggingface.co/Comfy-Org/Qwen-Image-Edit_ComfyUI/resolve/main/split_files/diffusion_models/qwen_image_edit_2509_fp8mixed.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-38 filename=qwen_image_edit_2511_fp8_e4m3fn.safetensors; size=20.4GB; type=unet; destination=diffusion_models; repository=drbaph/Qwen-Image-Edit-2511-FP8; host=huggingface.co; url=https://huggingface.co/drbaph/Qwen-Image-Edit-2511-FP8/resolve/main/qwen_image_edit_2511_fp8_e4m3fn.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-39 filename=qwen_image_edit_2511_fp8mixed.safetensors; size=19.12GB; type=unet; destination=diffusion_models; repository=Comfy-Org/Qwen-Image-Edit_ComfyUI; host=huggingface.co; url=https://huggingface.co/Comfy-Org/Qwen-Image-Edit_ComfyUI/resolve/main/split_files/diffusion_models/qwen_image_edit_2511_fp8mixed.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-40 filename=qwen_image_edit_fp8_e4m3fn.safetensors; size=10.7GB; type=unet; destination=diffusion_models; repository=Comfy-Org/Qwen-Image-Edit_ComfyUI; host=huggingface.co; url=https://huggingface.co/Comfy-Org/Qwen-Image-Edit_ComfyUI/resolve/main/split_files/diffusion_models/qwen_image_edit_fp8_e4m3fn.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-41 filename=qwen_image_fp8_e4m3fn.safetensors; size=10.7GB; type=unet; destination=diffusion_models; repository=Comfy-Org/Qwen-Image_ComfyUI; host=huggingface.co; url=https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/diffusion_models/qwen_image_fp8_e4m3fn.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-42 filename=qwen3_4b_fp8_scaled.safetensors; size=4.11GB; type=clip; destination=text_encoders; repository=jiangchengchengNLP/qwen3-4b-fp8-scaled; host=huggingface.co; url=https://huggingface.co/jiangchengchengNLP/qwen3-4b-fp8-scaled/resolve/main/qwen3_4b_fp8_scaled.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-43 filename=sd3_medium_incl_clips_t5xxlfp8.safetensors; size=5.58GB; type=checkpoint; destination=checkpoints; repository=stabilityai/stable-diffusion-3-medium; host=huggingface.co; url=https://huggingface.co/stabilityai/stable-diffusion-3-medium/resolve/main/sd3_medium_incl_clips_t5xxlfp8.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-44 filename=sd3.5_large_fp8_scaled.safetensors; size=8.3GB; type=checkpoint; destination=checkpoints; repository=stabilityai/stable-diffusion-3.5-large; host=huggingface.co; url=https://huggingface.co/stabilityai/stable-diffusion-3.5-large/resolve/main/sd3.5_large_fp8_scaled.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-45 filename=seedvr2_ema_3b_fp8_e4m3fn.safetensors; size=3.39GB; type=unet; destination=diffusion_models; repository=numz/SeedVR2_comfyUI; host=huggingface.co; url=https://huggingface.co/numz/SeedVR2_comfyUI/resolve/main/seedvr2_ema_3b_fp8_e4m3fn.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-46 filename=seedvr2_ema_7b_fp8_e4m3fn.safetensors; size=8.24GB; type=unet; destination=diffusion_models; repository=numz/SeedVR2_comfyUI; host=huggingface.co; url=https://huggingface.co/numz/SeedVR2_comfyUI/resolve/main/seedvr2_ema_7b_fp8_e4m3fn.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-47 filename=sulphur_dev_fp8mixed.safetensors; size=29.2GB; type=checkpoint; destination=checkpoints; repository=SulphurAI/Sulphur-2-base; host=huggingface.co; url=https://huggingface.co/SulphurAI/Sulphur-2-base/resolve/main/sulphur_dev_fp8mixed.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-48 filename=t5xxl_fp16.safetensors; size=9.79GB; type=clip; destination=clip; repository=comfyanonymous/flux_text_encoders; host=huggingface.co; url=https://huggingface.co/comfyanonymous/flux_text_encoders/resolve/main/t5xxl_fp16.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-49 filename=ultra_ae.safetensors; size=335MB; type=vae; destination=vae; repository=Owen777/UltraFlux-v1; host=huggingface.co; url=https://huggingface.co/Owen777/UltraFlux-v1/resolve/main/vae/diffusion_pytorch_model.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-50 filename=UltraFlux-v1_model.safetensors; size=335MB; type=vae; destination=vae; repository=Owen777/UltraFlux-v1; host=huggingface.co; url=https://huggingface.co/Owen777/UltraFlux-v1/resolve/main/vae/diffusion_pytorch_model.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-51 filename=umt5_xxl_fp8_e4m3fn_scaled.safetensors; size=4.9GB; type=clip; destination=text_encoders; repository=Comfy-Org/Wan_2.1_ComfyUI_repackaged; host=huggingface.co; url=https://huggingface.co/Comfy-Org/Wan_2.1_ComfyUI_repackaged/resolve/main/split_files/text_encoders/umt5_xxl_fp8_e4m3fn_scaled.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-52 filename=Wan2_2-Animate-14B_fp8_e4m3fn_scaled_KJ.safetensors; size=18.4GB; type=unet; destination=diffusion_models; repository=Kijai/WanVideo_comfy_fp8_scaled; host=huggingface.co; url=https://huggingface.co/Kijai/WanVideo_comfy_fp8_scaled/resolve/main/Wan22Animate/Wan2_2-Animate-14B_fp8_e4m3fn_scaled_KJ.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-53 filename=Wan2_2-Animate-14B_fp8_scaled_e4m3fn_KJ_v2.safetensors; size=17.3GB; type=unet; destination=diffusion_models; repository=Kijai/WanVideo_comfy_fp8_scaled; host=huggingface.co; url=https://huggingface.co/Kijai/WanVideo_comfy_fp8_scaled/resolve/main/Wan22Animate/Wan2_2-Animate-14B_fp8_scaled_e4m3fn_KJ_v2.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-54 filename=wan2.1_t2v_14B_fp8_e4m3fn.safetensors; size=14.3GB; type=unet; destination=diffusion_models; repository=Comfy-Org/Wan_2.1_ComfyUI_repackaged; host=huggingface.co; url=https://huggingface.co/Comfy-Org/Wan_2.1_ComfyUI_repackaged/resolve/main/split_files/diffusion_models/wan2.1_t2v_14B_fp8_e4m3fn.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-55 filename=wan2.2_fun_camera_high_noise_14B_fp8_scaled.safetensors; size=14.25GB; type=unet; destination=diffusion_models; repository=Comfy-Org/Wan_2.2_ComfyUI_Repackaged; host=huggingface.co; url=https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/resolve/main/split_files/diffusion_models/wan2.2_fun_camera_high_noise_14B_fp8_scaled.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-56 filename=wan2.2_fun_control_high_noise_14B_fp8_scaled.safetensors; size=13.31GB; type=unet; destination=diffusion_models; repository=Comfy-Org/Wan_2.2_ComfyUI_Repackaged; host=huggingface.co; url=https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/resolve/main/split_files/diffusion_models/wan2.2_fun_control_high_noise_14B_fp8_scaled.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-57 filename=wan2.2_i2v_high_noise_14B_fp8_scaled.safetensors; size=13.31GB; type=unet; destination=diffusion_models; repository=Comfy-Org/Wan_2.2_ComfyUI_Repackaged; host=huggingface.co; url=https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/resolve/main/split_files/diffusion_models/wan2.2_i2v_high_noise_14B_fp8_scaled.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-58 filename=wan2.2_i2v_low_noise_14B_fp8_scaled.safetensors; size=13.31GB; type=unet; destination=diffusion_models; repository=Comfy-Org/Wan_2.2_ComfyUI_Repackaged; host=huggingface.co; url=https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/resolve/main/split_files/diffusion_models/wan2.2_i2v_low_noise_14B_fp8_scaled.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-59 filename=wan2.2_t2v_high_noise_14B_fp8_scaled.safetensors; size=14.2GB; type=unet; destination=diffusion_models; repository=Comfy-Org/Wan_2.2_ComfyUI_Repackaged; host=huggingface.co; url=https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/resolve/main/split_files/diffusion_models/wan2.2_t2v_high_noise_14B_fp8_scaled.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-60 filename=wan2.2_t2v_low_noise_14B_fp8_scaled.safetensors; size=14.2GB; type=unet; destination=diffusion_models; repository=Comfy-Org/Wan_2.2_ComfyUI_Repackaged; host=huggingface.co; url=https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/resolve/main/split_files/diffusion_models/wan2.2_t2v_low_noise_14B_fp8_scaled.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-61 filename=z_image_turbo_fp8.safetensors; size=6.15GB; type=unet; destination=diffusion_models; repository=Comfy-Org/z_image_turbo; host=huggingface.co; url=https://huggingface.co/Comfy-Org/z_image_turbo/resolve/main/split_files/diffusion_models/z_image_turbo_fp8.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-62 filename=z-anime-base-fp8.safetensors; size=6.15GB; type=unet; destination=diffusion_models; repository=SeeSee21/Z-Anime; host=huggingface.co; url=https://huggingface.co/SeeSee21/Z-Anime/resolve/main/diffusion_models/z-anime-base-fp8.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-63 filename=z-anime-distill-4step-fp8.safetensors; size=6.15GB; type=unet; destination=diffusion_models; repository=SeeSee21/Z-Anime; host=huggingface.co; url=https://huggingface.co/SeeSee21/Z-Anime/resolve/main/diffusion_models/z-anime-distill-4step-fp8.safetensors
- flux2-klein-9b-fp8__flux2-klein-4b asset-64 filename=z-anime-distill-8step-fp8.safetensors; size=6.15GB; type=unet; destination=diffusion_models; repository=SeeSee21/Z-Anime; host=huggingface.co; url=https://huggingface.co/SeeSee21/Z-Anime/resolve/main/diffusion_models/z-anime-distill-8step-fp8.safetensors
Paired source evidence
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=id; value=flux2-klein-9b-fp8
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=name; value=FLUX.2 Klein 9B FP8
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=displayName; value=FLUX.2 Klein
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=sizeVariant; value=9B FP8 (Recommended)
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=templateName; value=comfyui-flux2
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=dockerImageBaseName; value=comfyui
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=dockerImageTag; value=flux2
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=modelSize; value=medium
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=popularity; value=11
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=modelGroup; value=flux2-klein
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=groupPopularity; value=99
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=h1; value=Deploy FLUX.2 Klein
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=aboutBlurb; value=FLUX.2 Klein is the fastest model in the Flux family; The 9B FP8 variant delivers sub-second generation in 4 steps, while the 4B model is fully open source under Apache 2.0 and runs on consumer GPUs;
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=useCases; value=Sub-second image generation, Real-time AI applications, Consumer GPU deployment, Open-source commercial use (Apache 2.0)
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=relatedGroups; value=flux-2, z-image-turbo, flux
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=configKeys; value=flux2-klein-9b-fp8, flux2-klein-4b, flux2-klein-9b
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=workflowUrl; value=https://docs.comfy.org/tutorials/flux/flux-2-klein
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=defaultConfig.templateName; value=comfyui-flux
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=defaultConfig.dockerImageBaseName; value=comfyui-flux
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=defaultConfig.containerPort; value=8188
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=defaultConfig.volumeMountPath; value=/data
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=defaultConfig.modelSize; value=medium
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=resolvedConfig.templateName; value=comfyui-flux2
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=resolvedConfig.dockerImageBaseName; value=comfyui
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=resolvedConfig.containerPort; value=8188
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=resolvedConfig.volumeMountPath; value=/data
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=resolvedConfig.modelSize; value=medium
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=recommendedInstance.id; value=gpu-l4
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=recommendedInstance.name; value=GPU Efficient (L4)
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=recommendedInstance.suitableFor; value=7B-13B LLMs, SDXL Images, Power Efficient
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=recommendedInstance.vCPU; value=12
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=recommendedInstance.ram; value=50
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=recommendedInstance.disk; value=200
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=recommendedInstance.gpuType; value=NVIDIA L4
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=recommendedInstance.gpuCount; value=1
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=recommendedInstance.type; value=gpu-efficient
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=recommendedInstance.deployCost; value=0.39
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=slug; value=flux-2-klein
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-9b-fp8; field=modelType; value=image
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=id; value=flux2-klein-4b
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=name; value=FLUX.2 Klein 4B
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=displayName; value=FLUX.2 Klein
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=sizeVariant; value=4B (Apache 2.0)
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=templateName; value=comfyui-flux2
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=dockerImageBaseName; value=comfyui
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=dockerImageTag; value=flux2
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=modelSize; value=small
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=popularity; value=10
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=modelGroup; value=flux2-klein
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=groupPopularity; value=99
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=h1; value=Deploy FLUX.2 Klein
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=aboutBlurb; value=FLUX.2 Klein is the fastest model in the Flux family; The 9B FP8 variant delivers sub-second generation in 4 steps, while the 4B model is fully open source under Apache 2.0 and runs on consumer GPUs;
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=useCases; value=Sub-second image generation, Real-time AI applications, Consumer GPU deployment, Open-source commercial use (Apache 2.0)
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=relatedGroups; value=flux-2, z-image-turbo, flux
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=configKeys; value=flux2-klein-9b-fp8, flux2-klein-4b, flux2-klein-9b
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=workflowUrl; value=https://docs.comfy.org/tutorials/flux/flux-2-klein
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=defaultConfig.templateName; value=comfyui-flux
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=defaultConfig.dockerImageBaseName; value=comfyui-flux
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=defaultConfig.containerPort; value=8188
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=defaultConfig.volumeMountPath; value=/data
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=defaultConfig.modelSize; value=medium
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=resolvedConfig.templateName; value=comfyui-flux2
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=resolvedConfig.dockerImageBaseName; value=comfyui
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=resolvedConfig.containerPort; value=8188
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=resolvedConfig.volumeMountPath; value=/data
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=resolvedConfig.modelSize; value=small
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=recommendedInstance.id; value=gpu-l4
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=recommendedInstance.name; value=GPU Efficient (L4)
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=recommendedInstance.suitableFor; value=7B-13B LLMs, SDXL Images, Power Efficient
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=recommendedInstance.vCPU; value=12
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=recommendedInstance.ram; value=50
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=recommendedInstance.disk; value=200
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=recommendedInstance.gpuType; value=NVIDIA L4
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=recommendedInstance.gpuCount; value=1
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=recommendedInstance.type; value=gpu-efficient
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=recommendedInstance.deployCost; value=0.39
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=slug; value=flux-2-klein
- flux2-klein-9b-fp8__flux2-klein-4b source-model=flux2-klein-4b; field=modelType; value=image
Pair-scoped catalog context
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-1 nearby=flux2-klein-9b-fp8; relation=selected-model-a; group=flux2-klein; size=9B FP8 (Recommended)/medium; instance=gpu-l4; hourly=0.39; monthly=280.8; delta-vs-flux2-klein-9b-fp8=0; delta-vs-flux2-klein-4b=0; template=comfyui-flux2; docker=comfyui; template-match-a=yes; template-match-b=yes; term-overlap-a=14; term-overlap-b=13; source=flux2-klein-9b-fp8; description=Fastest Flux model; Sub-second generation, 4-step distilled; Runs on RTX 4090;; useCases=Sub-second image generation; Real-time AI applications; Consumer GPU deployment; Open-source commercial use (Apache 2.0); config=id=flux2-klein-9b-fp8 | name=FLUX.2 Klein 9B FP8 | displayName=FLUX.2 Klein | sizeVariant=9B FP8 (Recommended) | templateName=comfyui-flux2 | dockerImageBaseName=comfyui | dockerImageTag=flux2 | workflowUrl=https://docs.comfy.org/tutorials/flux/flux-2-klein | modelSize=medium | popularity=11 | modelGroup=flux2-klein | groupPopularity=99 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-flux2 | resolvedConfig.dockerImageBaseName=comfyui | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=medium | recommendedInstance.id=gpu-l4 | recommendedInstance.name=GPU Efficient (L4) | recommendedInstance.suitableFor=7B-13B LLMs, SDXL Images, Power Efficient | recommendedInstance.vCPU=12 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA L4 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-efficient | recommendedInstance.deployCost=0.39; seo=slug=flux-2-klein | modelType=image | h1=Deploy FLUX.2 Klein | aboutBlurb=FLUX.2 Klein is the fastest model in the Flux family; The 9B FP8 variant delivers sub-second generation in 4 steps, while the 4B model is fully open source under Apache 2.0 and runs on consumer GPUs; | useCases=Sub-second image generation, Real-time AI applications, Consumer GPU deployment, Open-source commercial use (Apache 2.0) | relatedGroups=flux-2, z-image-turbo, flux | configKeys=flux2-klein-9b-fp8, flux2-klein-4b, flux2-klein-9b; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-2 nearby=flux2-klein-4b; relation=selected-model-b; group=flux2-klein; size=4B (Apache 2.0)/small; instance=gpu-l4; hourly=0.39; monthly=280.8; delta-vs-flux2-klein-9b-fp8=0; delta-vs-flux2-klein-4b=0; template=comfyui-flux2; docker=comfyui; template-match-a=yes; template-match-b=yes; term-overlap-a=13; term-overlap-b=14; source=flux2-klein-4b; description=Fully open source (Apache 2.0); Fast generation on RTX 3090/4070; ~13GB VRAM;; useCases=Sub-second image generation; Real-time AI applications; Consumer GPU deployment; Open-source commercial use (Apache 2.0); config=id=flux2-klein-4b | name=FLUX.2 Klein 4B | displayName=FLUX.2 Klein | sizeVariant=4B (Apache 2.0) | templateName=comfyui-flux2 | dockerImageBaseName=comfyui | dockerImageTag=flux2 | workflowUrl=https://docs.comfy.org/tutorials/flux/flux-2-klein | modelSize=small | popularity=10 | modelGroup=flux2-klein | groupPopularity=99 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-flux2 | resolvedConfig.dockerImageBaseName=comfyui | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=small | recommendedInstance.id=gpu-l4 | recommendedInstance.name=GPU Efficient (L4) | recommendedInstance.suitableFor=7B-13B LLMs, SDXL Images, Power Efficient | recommendedInstance.vCPU=12 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA L4 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-efficient | recommendedInstance.deployCost=0.39; seo=slug=flux-2-klein | modelType=image | h1=Deploy FLUX.2 Klein | aboutBlurb=FLUX.2 Klein is the fastest model in the Flux family; The 9B FP8 variant delivers sub-second generation in 4 steps, while the 4B model is fully open source under Apache 2.0 and runs on consumer GPUs; | useCases=Sub-second image generation, Real-time AI applications, Consumer GPU deployment, Open-source commercial use (Apache 2.0) | relatedGroups=flux-2, z-image-turbo, flux | configKeys=flux2-klein-9b-fp8, flux2-klein-4b, flux2-klein-9b; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-3 nearby=flux2-klein-9b; relation=same-group-as-flux2-klein-9b-fp8; group=flux2-klein; size=9B Base (Undistilled)/medium; instance=gpu-l4; hourly=0.39; monthly=280.8; delta-vs-flux2-klein-9b-fp8=0; delta-vs-flux2-klein-4b=0; template=comfyui-flux2; docker=comfyui; template-match-a=yes; template-match-b=yes; term-overlap-a=13; term-overlap-b=13; source=flux2-klein-9b; description=Base model for fine-tuning and LoRA training; More steps but higher quality ceiling;; useCases=Sub-second image generation; Real-time AI applications; Consumer GPU deployment; Open-source commercial use (Apache 2.0); config=id=flux2-klein-9b | name=FLUX.2 Klein 9B Base | displayName=FLUX.2 Klein | sizeVariant=9B Base (Undistilled) | templateName=comfyui-flux2 | dockerImageBaseName=comfyui | dockerImageTag=flux2 | workflowUrl=https://docs.comfy.org/tutorials/flux/flux-2-klein | modelSize=medium | popularity=6 | modelGroup=flux2-klein | groupPopularity=99 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-flux2 | resolvedConfig.dockerImageBaseName=comfyui | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=medium | recommendedInstance.id=gpu-l4 | recommendedInstance.name=GPU Efficient (L4) | recommendedInstance.suitableFor=7B-13B LLMs, SDXL Images, Power Efficient | recommendedInstance.vCPU=12 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA L4 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-efficient | recommendedInstance.deployCost=0.39; seo=slug=flux-2-klein | modelType=image | h1=Deploy FLUX.2 Klein | aboutBlurb=FLUX.2 Klein is the fastest model in the Flux family; The 9B FP8 variant delivers sub-second generation in 4 steps, while the 4B model is fully open source under Apache 2.0 and runs on consumer GPUs; | useCases=Sub-second image generation, Real-time AI applications, Consumer GPU deployment, Open-source commercial use (Apache 2.0) | relatedGroups=flux-2, z-image-turbo, flux | configKeys=flux2-klein-9b-fp8, flux2-klein-4b, flux2-klein-9b; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-4 nearby=flux2-dev-q4; relation=same-deployment-tier; group=flux2; size=Dev Q4 (Low VRAM)/medium; instance=gpu-l4; hourly=0.39; monthly=280.8; delta-vs-flux2-klein-9b-fp8=0; delta-vs-flux2-klein-4b=0; template=comfyui-flux2; docker=comfyui; template-match-a=yes; template-match-b=yes; term-overlap-a=4; term-overlap-b=4; source=flux2-dev-q4; description=Quantized for consumer GPUs; 20GB model, runs on RTX 3090/4080;; useCases=Multi-image editing workflows; Professional image generation; Fine-grained image control; Production-quality outputs; config=id=flux2-dev-q4 | name=FLUX.2 Dev Q4 GGUF | displayName=FLUX.2 | sizeVariant=Dev Q4 (Low VRAM) | templateName=comfyui-flux2 | dockerImageBaseName=comfyui | dockerImageTag=flux2 | workflowUrl=https://comfyanonymous.github.io/ComfyUI_examples/flux2/ | modelSize=medium | popularity=7 | modelGroup=flux2 | groupPopularity=98 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-flux2 | resolvedConfig.dockerImageBaseName=comfyui | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=medium | recommendedInstance.id=gpu-l4 | recommendedInstance.name=GPU Efficient (L4) | recommendedInstance.suitableFor=7B-13B LLMs, SDXL Images, Power Efficient | recommendedInstance.vCPU=12 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA L4 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-efficient | recommendedInstance.deployCost=0.39; seo=slug=flux-2 | modelType=image | h1=Deploy FLUX.2 | aboutBlurb=FLUX.2 is Black Forest Labs' 32B parameter model with multi-image editing capabilities; The Dev FP8 variant runs on RTX 4090/5090, while the full model requires 48GB+ VRAM for maximum quality; | useCases=Multi-image editing workflows, Professional image generation, Fine-grained image control, Production-quality outputs | relatedGroups=flux-2-klein, flux, hidream-i1 | configKeys=flux2-dev-fp8, flux2-dev, flux2-dev-q4; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-5 nearby=flux-kontext-dev-fp8; relation=same-deployment-tier; group=flux-kontext; size=Dev FP8 (Character Consistency)/medium; instance=gpu-l4; hourly=0.39; monthly=280.8; delta-vs-flux2-klein-9b-fp8=0; delta-vs-flux2-klein-4b=0; template=comfyui-flux; docker=comfyui-flux; template-match-a=no; template-match-b=no; term-overlap-a=1; term-overlap-b=1; source=flux-kontext-dev-fp8; description=Character consistency via reference images; Upload a face/character photo and generate new scenes preserving identity; No custom nodes needed;; useCases=Character-consistent social media content; Brand mascot in different scenes; Visual storytelling and comics; Character design exploration; Consistent avatar generation; config=id=flux-kontext-dev-fp8 | name=Flux Kontext Dev FP8 | displayName=Flux Kontext | sizeVariant=Dev FP8 (Character Consistency) | templateName=comfyui-flux | dockerImageBaseName=comfyui-flux | dockerImageTag=flux-dev | workflowUrl=https://docs.comfy.org/tutorials/flux/flux-1-kontext-dev | modelSize=medium | popularity=11 | modelGroup=flux-kontext | groupPopularity=97 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-flux | resolvedConfig.dockerImageBaseName=comfyui-flux | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=medium | recommendedInstance.id=gpu-l4 | recommendedInstance.name=GPU Efficient (L4) | recommendedInstance.suitableFor=7B-13B LLMs, SDXL Images, Power Efficient | recommendedInstance.vCPU=12 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA L4 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-efficient | recommendedInstance.deployCost=0.39; seo=slug=flux-kontext | modelType=image | h1=Deploy Flux Kontext | aboutBlurb=Flux Kontext Dev is a character consistency model from Black Forest Labs; Upload a reference photo and generate new scenes that preserve the character's identity — face, clothing, style, and features; Uses native in-context conditioning (no adapters or custom nodes), achieving 98% identity retention across generations; The FP8 variant runs on L4 (24GB VRAM) at $0.53/hr; | useCases=Character-consistent social media content, Brand mascot in different scenes, Visual storytelling and comics, Character design exploration, Consistent avatar generation | relatedGroups=flux, flux-2, hidream-i1 | configKeys=flux-kontext-dev-fp8; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-6 nearby=z-image-turbo; relation=same-deployment-tier; group=z-image; size=Turbo/medium; instance=gpu-l4; hourly=0.39; monthly=280.8; delta-vs-flux2-klein-9b-fp8=0; delta-vs-flux2-klein-4b=0; template=comfyui-zimage; docker=comfyui; template-match-a=no; template-match-b=no; term-overlap-a=5; term-overlap-b=5; source=z-image-turbo; description=Fast distilled model; Sub-second generation on H800;; useCases=Real-time image generation; Rapid prototyping and iteration; Interactive AI applications; Batch image processing; config=id=z-image-turbo | name=Z Image Turbo | displayName=Z Image | sizeVariant=Turbo | templateName=comfyui-zimage | dockerImageBaseName=comfyui | dockerImageTag=z-image-turbo | workflowUrl=https://comfyanonymous.github.io/ComfyUI_examples/z_image/ | modelSize=medium | popularity=10 | modelGroup=z-image | groupPopularity=100 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-zimage | resolvedConfig.dockerImageBaseName=comfyui | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=medium | recommendedInstance.id=gpu-l4 | recommendedInstance.name=GPU Efficient (L4) | recommendedInstance.suitableFor=7B-13B LLMs, SDXL Images, Power Efficient | recommendedInstance.vCPU=12 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA L4 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-efficient | recommendedInstance.deployCost=0.39; seo=slug=z-image-turbo | modelType=image | h1=Deploy Z Image Turbo | aboutBlurb=Z Image Turbo is a fast distilled image generation model from Alibaba Tongyi; It achieves sub-second generation times on high-end GPUs, making it ideal for real-time applications and rapid iteration; | useCases=Real-time image generation, Rapid prototyping and iteration, Interactive AI applications, Batch image processing | relatedGroups=flux-2-klein, qwen-image-2512, flux | configKeys=z-image-turbo; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-7 nearby=flux-dev; relation=same-deployment-tier; group=flux; size=Dev (Quality)/medium; instance=gpu-l4; hourly=0.39; monthly=280.8; delta-vs-flux2-klein-9b-fp8=0; delta-vs-flux2-klein-4b=0; template=comfyui-flux; docker=comfyui-flux; template-match-a=no; template-match-b=no; term-overlap-a=2; term-overlap-b=2; source=flux-dev; description=Best quality; Excellent text rendering and composition;; useCases=High-quality image generation; Text rendering in images; Creative art and illustration; Brand asset creation; config=id=flux-dev | name=Flux Dev | displayName=Flux | sizeVariant=Dev (Quality) | templateName=comfyui-flux | dockerImageBaseName=comfyui-flux | dockerImageTag=flux-dev | workflowUrl=https://comfyanonymous.github.io/ComfyUI_examples/flux/flux-dev-example.json | modelSize=medium | popularity=9 | modelGroup=flux | groupPopularity=95 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-flux | resolvedConfig.dockerImageBaseName=comfyui-flux | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=medium | recommendedInstance.id=gpu-l4 | recommendedInstance.name=GPU Efficient (L4) | recommendedInstance.suitableFor=7B-13B LLMs, SDXL Images, Power Efficient | recommendedInstance.vCPU=12 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA L4 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-efficient | recommendedInstance.deployCost=0.39; seo=slug=flux | modelType=image | h1=Deploy Flux | aboutBlurb=Flux is Black Forest Labs' flagship image generation family; Flux Dev delivers the best quality with excellent text rendering, while Flux Schnell offers fast 4-step generation; Flux Krea provides a natural aesthetic without the typical AI look; | useCases=High-quality image generation, Text rendering in images, Creative art and illustration, Brand asset creation | relatedGroups=flux-kontext, flux-2, flux-2-klein, qwen-image-2512 | configKeys=flux-dev, flux-schnell, flux-krea-dev; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-8 nearby=z-anime; relation=same-deployment-tier; group=z-image; size=Distill 4-step/medium; instance=gpu-l4; hourly=0.39; monthly=280.8; delta-vs-flux2-klein-9b-fp8=0; delta-vs-flux2-klein-4b=0; template=comfyui-zimage; docker=comfyui; template-match-a=no; template-match-b=no; term-overlap-a=1; term-overlap-b=1; source=z-anime; description=SeeSee21 anime fine-tune of Z-Image (6B, Apache-2.0); 4-step distill at fp8, ~3-4s per 1024x1024 image on an L4; Natural-language prompting (no booru tags);; useCases=; config=id=z-anime | name=Z-Anime (4-step distill) | displayName=Z Anime | sizeVariant=Distill 4-step | templateName=comfyui-zimage | dockerImageBaseName=comfyui | dockerImageTag=z-image-turbo | workflowUrl=https://huggingface.co/SeeSee21/Z-Anime/resolve/main/workflows/Z-Anime-Workflow-v1.json | modelSize=medium | popularity=9 | modelGroup=z-image | groupPopularity=100 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-zimage | resolvedConfig.dockerImageBaseName=comfyui | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=medium | recommendedInstance.id=gpu-l4 | recommendedInstance.name=GPU Efficient (L4) | recommendedInstance.suitableFor=7B-13B LLMs, SDXL Images, Power Efficient | recommendedInstance.vCPU=12 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA L4 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-efficient | recommendedInstance.deployCost=0.39; seo=; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-9 nearby=flux-schnell; relation=same-deployment-tier; group=flux; size=Schnell (Fast)/medium; instance=gpu-l4; hourly=0.39; monthly=280.8; delta-vs-flux2-klein-9b-fp8=0; delta-vs-flux2-klein-4b=0; template=comfyui-flux; docker=comfyui-flux; template-match-a=no; template-match-b=no; term-overlap-a=2; term-overlap-b=2; source=flux-schnell; description=Fast 4-step Flux; Good quality with quick generation;; useCases=High-quality image generation; Text rendering in images; Creative art and illustration; Brand asset creation; config=id=flux-schnell | name=Flux Schnell | displayName=Flux | sizeVariant=Schnell (Fast) | templateName=comfyui-flux | dockerImageBaseName=comfyui-flux | dockerImageTag=flux-dev | workflowUrl=https://comfyanonymous.github.io/ComfyUI_examples/flux/flux-schnell-example.json | modelSize=medium | popularity=8 | modelGroup=flux | groupPopularity=95 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-flux | resolvedConfig.dockerImageBaseName=comfyui-flux | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=medium | recommendedInstance.id=gpu-l4 | recommendedInstance.name=GPU Efficient (L4) | recommendedInstance.suitableFor=7B-13B LLMs, SDXL Images, Power Efficient | recommendedInstance.vCPU=12 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA L4 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-efficient | recommendedInstance.deployCost=0.39; seo=slug=flux | modelType=image | h1=Deploy Flux | aboutBlurb=Flux is Black Forest Labs' flagship image generation family; Flux Dev delivers the best quality with excellent text rendering, while Flux Schnell offers fast 4-step generation; Flux Krea provides a natural aesthetic without the typical AI look; | useCases=High-quality image generation, Text rendering in images, Creative art and illustration, Brand asset creation | relatedGroups=flux-kontext, flux-2, flux-2-klein, qwen-image-2512 | configKeys=flux-dev, flux-schnell, flux-krea-dev; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-10 nearby=flux2-dev-fp8; relation=same-type-neighbor; group=flux2; size=Dev FP8 (Balanced)/large; instance=gpu-a6000; hourly=0.49; monthly=352.8; delta-vs-flux2-klein-9b-fp8=0.1; delta-vs-flux2-klein-4b=0.1; template=comfyui-flux2; docker=comfyui; template-match-a=yes; template-match-b=yes; term-overlap-a=4; term-overlap-b=4; source=flux2-dev-fp8; description=32B model with multi-image editing; Runs on RTX 4090/5090; Best open-weight model;; useCases=Multi-image editing workflows; Professional image generation; Fine-grained image control; Production-quality outputs; config=id=flux2-dev-fp8 | name=FLUX.2 Dev FP8 | displayName=FLUX.2 | sizeVariant=Dev FP8 (Balanced) | templateName=comfyui-flux2 | dockerImageBaseName=comfyui | dockerImageTag=flux2 | workflowUrl=https://comfyanonymous.github.io/ComfyUI_examples/flux2/ | modelSize=large | popularity=10 | modelGroup=flux2 | groupPopularity=98 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-flux2 | resolvedConfig.dockerImageBaseName=comfyui | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=large | recommendedInstance.id=gpu-a6000 | recommendedInstance.name=GPU High VRAM (A6000) | recommendedInstance.suitableFor=30B+ LLMs (quantized 70B), Video Gen, 48GB VRAM | recommendedInstance.vCPU=9 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA RTX A6000 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-high-vram-value | recommendedInstance.deployCost=0.49; seo=slug=flux-2 | modelType=image | h1=Deploy FLUX.2 | aboutBlurb=FLUX.2 is Black Forest Labs' 32B parameter model with multi-image editing capabilities; The Dev FP8 variant runs on RTX 4090/5090, while the full model requires 48GB+ VRAM for maximum quality; | useCases=Multi-image editing workflows, Professional image generation, Fine-grained image control, Production-quality outputs | relatedGroups=flux-2-klein, flux, hidream-i1 | configKeys=flux2-dev-fp8, flux2-dev, flux2-dev-q4; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-11 nearby=stable-diffusion-xl; relation=same-deployment-tier; group=stable-diffusion; size=XL/medium; instance=gpu-l4; hourly=0.39; monthly=280.8; delta-vs-flux2-klein-9b-fp8=0; delta-vs-flux2-klein-4b=0; template=comfyui-dynamic; docker=comfyui; template-match-a=no; template-match-b=no; term-overlap-a=2; term-overlap-b=2; source=stable-diffusion-xl; description=Most fine-tunes and LoRAs available; 1024x1024;; useCases=LoRA and fine-tune deployment; Custom model training base; Established workflow integration; Community model ecosystem; config=id=stable-diffusion-xl | name=Stable Diffusion XL | displayName=Stable Diffusion | sizeVariant=XL | templateName=comfyui-dynamic | dockerImageBaseName=comfyui | dockerImageTag=dynamic | modelSize=medium | popularity=7 | modelGroup=stable-diffusion | groupPopularity=90 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-dynamic | resolvedConfig.dockerImageBaseName=comfyui | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=medium | recommendedInstance.id=gpu-l4 | recommendedInstance.name=GPU Efficient (L4) | recommendedInstance.suitableFor=7B-13B LLMs, SDXL Images, Power Efficient | recommendedInstance.vCPU=12 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA L4 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-efficient | recommendedInstance.deployCost=0.39; seo=slug=stable-diffusion | modelType=image | h1=Deploy Stable Diffusion | aboutBlurb=Stable Diffusion by Stability AI is the most widely adopted image generation family with the largest ecosystem of fine-tunes and LoRAs; SDXL generates 1024x1024 images, while SD 3.5 offers improved text rendering; | useCases=LoRA and fine-tune deployment, Custom model training base, Established workflow integration, Community model ecosystem | relatedGroups=flux, qwen-image-2512, z-image-turbo | configKeys=stable-diffusion-xl, stable-diffusion-35-large, stable-diffusion-35-medium, stable-diffusion-35-turbo, stable-diffusion-1.5; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-12 nearby=flux2-dev; relation=same-type-neighbor; group=flux2; size=Dev Full (64GB)/xlarge; instance=gpu-a100-80gb; hourly=1.39; monthly=1000.8; delta-vs-flux2-klein-9b-fp8=1; delta-vs-flux2-klein-4b=1; template=comfyui-flux2; docker=comfyui; template-match-a=yes; template-match-b=yes; term-overlap-a=4; term-overlap-b=4; source=flux2-dev; description=Full 32B model; Requires 48GB+ VRAM (H100); Maximum quality and fidelity;; useCases=Multi-image editing workflows; Professional image generation; Fine-grained image control; Production-quality outputs; config=id=flux2-dev | name=FLUX.2 Dev Full | displayName=FLUX.2 | sizeVariant=Dev Full (64GB) | templateName=comfyui-flux2 | dockerImageBaseName=comfyui | dockerImageTag=flux2 | workflowUrl=https://comfyanonymous.github.io/ComfyUI_examples/flux2/ | modelSize=xlarge | popularity=8 | modelGroup=flux2 | groupPopularity=98 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-flux2 | resolvedConfig.dockerImageBaseName=comfyui | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=xlarge | recommendedInstance.id=gpu-a100-80gb | recommendedInstance.name=GPU Very High VRAM (A100 80GB) | recommendedInstance.suitableFor=70B+ LLMs, Complex Workflows, Fine-tuning | recommendedInstance.vCPU=8 | recommendedInstance.ram=117 | recommendedInstance.disk=300 | recommendedInstance.gpuType=NVIDIA A100 80GB PCIe | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-very-high-vram | recommendedInstance.deployCost=1.39; seo=slug=flux-2 | modelType=image | h1=Deploy FLUX.2 | aboutBlurb=FLUX.2 is Black Forest Labs' 32B parameter model with multi-image editing capabilities; The Dev FP8 variant runs on RTX 4090/5090, while the full model requires 48GB+ VRAM for maximum quality; | useCases=Multi-image editing workflows, Professional image generation, Fine-grained image control, Production-quality outputs | relatedGroups=flux-2-klein, flux, hidream-i1 | configKeys=flux2-dev-fp8, flux2-dev, flux2-dev-q4; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-13 nearby=flux-krea-dev; relation=same-deployment-tier; group=flux; size=Krea (Natural)/medium; instance=gpu-l4; hourly=0.39; monthly=280.8; delta-vs-flux2-klein-9b-fp8=0; delta-vs-flux2-klein-4b=0; template=comfyui-flux; docker=comfyui-flux; template-match-a=no; template-match-b=no; term-overlap-a=2; term-overlap-b=2; source=flux-krea-dev; description=Natural aesthetic without typical AI look;; useCases=High-quality image generation; Text rendering in images; Creative art and illustration; Brand asset creation; config=id=flux-krea-dev | name=Flux Krea | displayName=Flux | sizeVariant=Krea (Natural) | templateName=comfyui-flux | dockerImageBaseName=comfyui-flux | dockerImageTag=flux-dev | workflowUrl=https://comfyanonymous.github.io/ComfyUI_examples/flux/flux-dev-example.json | modelSize=medium | popularity=5 | modelGroup=flux | groupPopularity=95 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-flux | resolvedConfig.dockerImageBaseName=comfyui-flux | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=medium | recommendedInstance.id=gpu-l4 | recommendedInstance.name=GPU Efficient (L4) | recommendedInstance.suitableFor=7B-13B LLMs, SDXL Images, Power Efficient | recommendedInstance.vCPU=12 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA L4 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-efficient | recommendedInstance.deployCost=0.39; seo=slug=flux | modelType=image | h1=Deploy Flux | aboutBlurb=Flux is Black Forest Labs' flagship image generation family; Flux Dev delivers the best quality with excellent text rendering, while Flux Schnell offers fast 4-step generation; Flux Krea provides a natural aesthetic without the typical AI look; | useCases=High-quality image generation, Text rendering in images, Creative art and illustration, Brand asset creation | relatedGroups=flux-kontext, flux-2, flux-2-klein, qwen-image-2512 | configKeys=flux-dev, flux-schnell, flux-krea-dev; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-14 nearby=stable-diffusion-35-medium; relation=same-deployment-tier; group=stable-diffusion; size=Medium/medium; instance=gpu-l4; hourly=0.39; monthly=280.8; delta-vs-flux2-klein-9b-fp8=0; delta-vs-flux2-klein-4b=0; template=comfyui-dynamic; docker=comfyui; template-match-a=no; template-match-b=no; term-overlap-a=2; term-overlap-b=2; source=stable-diffusion-35-medium; description=Good balance of quality and speed;; useCases=LoRA and fine-tune deployment; Custom model training base; Established workflow integration; Community model ecosystem; config=id=stable-diffusion-35-medium | name=Stable Diffusion 3.5 Medium | displayName=Stable Diffusion 3.5 | sizeVariant=Medium | templateName=comfyui-dynamic | dockerImageBaseName=comfyui | dockerImageTag=dynamic | modelSize=medium | popularity=4 | modelGroup=stable-diffusion | groupPopularity=90 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-dynamic | resolvedConfig.dockerImageBaseName=comfyui | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=medium | recommendedInstance.id=gpu-l4 | recommendedInstance.name=GPU Efficient (L4) | recommendedInstance.suitableFor=7B-13B LLMs, SDXL Images, Power Efficient | recommendedInstance.vCPU=12 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA L4 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-efficient | recommendedInstance.deployCost=0.39; seo=slug=stable-diffusion | modelType=image | h1=Deploy Stable Diffusion | aboutBlurb=Stable Diffusion by Stability AI is the most widely adopted image generation family with the largest ecosystem of fine-tunes and LoRAs; SDXL generates 1024x1024 images, while SD 3.5 offers improved text rendering; | useCases=LoRA and fine-tune deployment, Custom model training base, Established workflow integration, Community model ecosystem | relatedGroups=flux, qwen-image-2512, z-image-turbo | configKeys=stable-diffusion-xl, stable-diffusion-35-large, stable-diffusion-35-medium, stable-diffusion-35-turbo, stable-diffusion-1.5; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-15 nearby=stable-diffusion-35-turbo; relation=same-deployment-tier; group=stable-diffusion; size=Turbo/medium; instance=gpu-l4; hourly=0.39; monthly=280.8; delta-vs-flux2-klein-9b-fp8=0; delta-vs-flux2-klein-4b=0; template=comfyui-dynamic; docker=comfyui; template-match-a=no; template-match-b=no; term-overlap-a=2; term-overlap-b=2; source=stable-diffusion-35-turbo; description=Fast SD 3.5 for quick iterations;; useCases=LoRA and fine-tune deployment; Custom model training base; Established workflow integration; Community model ecosystem; config=id=stable-diffusion-35-turbo | name=Stable Diffusion 3.5 Turbo | displayName=Stable Diffusion 3.5 | sizeVariant=Turbo | templateName=comfyui-dynamic | dockerImageBaseName=comfyui | dockerImageTag=dynamic | modelSize=medium | popularity=3 | modelGroup=stable-diffusion | groupPopularity=90 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-dynamic | resolvedConfig.dockerImageBaseName=comfyui | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=medium | recommendedInstance.id=gpu-l4 | recommendedInstance.name=GPU Efficient (L4) | recommendedInstance.suitableFor=7B-13B LLMs, SDXL Images, Power Efficient | recommendedInstance.vCPU=12 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA L4 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-efficient | recommendedInstance.deployCost=0.39; seo=slug=stable-diffusion | modelType=image | h1=Deploy Stable Diffusion | aboutBlurb=Stable Diffusion by Stability AI is the most widely adopted image generation family with the largest ecosystem of fine-tunes and LoRAs; SDXL generates 1024x1024 images, while SD 3.5 offers improved text rendering; | useCases=LoRA and fine-tune deployment, Custom model training base, Established workflow integration, Community model ecosystem | relatedGroups=flux, qwen-image-2512, z-image-turbo | configKeys=stable-diffusion-xl, stable-diffusion-35-large, stable-diffusion-35-medium, stable-diffusion-35-turbo, stable-diffusion-1.5; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-16 nearby=stable-diffusion-1.5; relation=same-deployment-tier; group=stable-diffusion; size=1.5 (Legacy)/small; instance=gpu-l4; hourly=0.39; monthly=280.8; delta-vs-flux2-klein-9b-fp8=0; delta-vs-flux2-klein-4b=0; template=comfyui-simple-test; docker=comfyui; template-match-a=no; template-match-b=no; term-overlap-a=2; term-overlap-b=2; source=stable-diffusion-1.5; description=Legacy model; Fast and lightweight for testing;; useCases=LoRA and fine-tune deployment; Custom model training base; Established workflow integration; Community model ecosystem; config=id=stable-diffusion-1.5 | name=Stable Diffusion 1.5 | displayName=Stable Diffusion | sizeVariant=1.5 (Legacy) | templateName=comfyui-simple-test | dockerImageBaseName=comfyui | dockerImageTag=simple-test | workflowUrl=bundled://simple-test | modelSize=small | popularity=2 | modelGroup=stable-diffusion | groupPopularity=90 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-simple-test | resolvedConfig.dockerImageBaseName=comfyui | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=small | recommendedInstance.id=gpu-l4 | recommendedInstance.name=GPU Efficient (L4) | recommendedInstance.suitableFor=7B-13B LLMs, SDXL Images, Power Efficient | recommendedInstance.vCPU=12 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA L4 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-efficient | recommendedInstance.deployCost=0.39; seo=slug=stable-diffusion | modelType=image | h1=Deploy Stable Diffusion | aboutBlurb=Stable Diffusion by Stability AI is the most widely adopted image generation family with the largest ecosystem of fine-tunes and LoRAs; SDXL generates 1024x1024 images, while SD 3.5 offers improved text rendering; | useCases=LoRA and fine-tune deployment, Custom model training base, Established workflow integration, Community model ecosystem | relatedGroups=flux, qwen-image-2512, z-image-turbo | configKeys=stable-diffusion-xl, stable-diffusion-35-large, stable-diffusion-35-medium, stable-diffusion-35-turbo, stable-diffusion-1.5; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-17 nearby=custom-comfyui; relation=same-deployment-tier; group=custom-comfyui; size=medium/medium; instance=gpu-l4; hourly=0.39; monthly=280.8; delta-vs-flux2-klein-9b-fp8=0; delta-vs-flux2-klein-4b=0; template=comfyui-flux; docker=comfyui-flux; template-match-a=no; template-match-b=no; term-overlap-a=0; term-overlap-b=0; source=custom-comfyui; description=Deploy ComfyUI with pre-saved workflows or custom node setups; useCases=; config=id=custom-comfyui | name=Custom ComfyUI Workflow | templateName=comfyui-flux | dockerImageBaseName=comfyui-flux | modelSize=medium | isCustom=true | customType=comfyui-workflow | presetWorkflows.0.name=Text-to-Image Basic | presetWorkflows.0.description=Simple text-to-image generation workflow | presetWorkflows.0.url=https://comfyanonymous.github.io/ComfyUI_examples/basic/text_to_image_basic.json | presetWorkflows.1.name=Image-to-Image | presetWorkflows.1.description=Image-to-image transformation workflow | presetWorkflows.1.url=https://comfyanonymous.github.io/ComfyUI_examples/basic/img2img_basic.json | presetWorkflows.2.name=Inpainting | presetWorkflows.2.description=Image inpainting and editing workflow | presetWorkflows.2.url=https://comfyanonymous.github.io/ComfyUI_examples/basic/inpainting_basic.json | presetWorkflows.3.name=ControlNet | presetWorkflows.3.description=Controlled image generation with ControlNet | presetWorkflows.3.url=https://comfyanonymous.github.io/ComfyUI_examples/controlnet/controlnet_basic.json | presetWorkflows.4.name=Custom Upload | presetWorkflows.4.description=Upload your own workflow JSON file | presetWorkflows.4.url=custom-upload | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-flux | resolvedConfig.dockerImageBaseName=comfyui-flux | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=medium | recommendedInstance.id=gpu-l4 | recommendedInstance.name=GPU Efficient (L4) | recommendedInstance.suitableFor=7B-13B LLMs, SDXL Images, Power Efficient | recommendedInstance.vCPU=12; seo=; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-18 nearby=custom-workflow; relation=same-deployment-tier; group=custom-workflow; size=medium/medium; instance=gpu-l4; hourly=0.39; monthly=280.8; delta-vs-flux2-klein-9b-fp8=0; delta-vs-flux2-klein-4b=0; template=comfyui-dynamic; docker=comfyui; template-match-a=no; template-match-b=no; term-overlap-a=1; term-overlap-b=1; source=custom-workflow; description=User-uploaded ComfyUI workflow with automatic custom node installation; useCases=; config=id=custom-workflow | name=Custom Workflow | templateName=comfyui-dynamic | dockerImageBaseName=comfyui | dockerImageTag=dynamic | modelSize=medium | isCustom=true | customType=workflow-upload | allowDynamicImageOverride=true | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-dynamic | resolvedConfig.dockerImageBaseName=comfyui | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=medium | recommendedInstance.id=gpu-l4 | recommendedInstance.name=GPU Efficient (L4) | recommendedInstance.suitableFor=7B-13B LLMs, SDXL Images, Power Efficient | recommendedInstance.vCPU=12 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA L4 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-efficient | recommendedInstance.deployCost=0.39; seo=; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-19 nearby=qwen-image-2512; relation=same-type-neighbor; group=qwen-image; size=2512 (Latest)/large; instance=gpu-a6000; hourly=0.49; monthly=352.8; delta-vs-flux2-klein-9b-fp8=0.1; delta-vs-flux2-klein-4b=0.1; template=comfyui-qwen-image-2512; docker=comfyui; template-match-a=no; template-match-b=no; term-overlap-a=2; term-overlap-b=2; source=qwen-image-2512; description=#1 open-source model; Superior text rendering, realistic humans;; useCases=Marketing and advertising visuals; Product photography; Text-heavy graphic design; Realistic portrait generation; config=id=qwen-image-2512 | name=Qwen-Image-2512 | displayName=Qwen Image | sizeVariant=2512 (Latest) | templateName=comfyui-qwen-image-2512 | dockerImageBaseName=comfyui | dockerImageTag=qwen-image-2512 | workflowUrl=https://comfyanonymous.github.io/ComfyUI_examples/qwen_image/ | modelSize=large | popularity=11 | modelGroup=qwen-image | groupPopularity=105 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-qwen-image-2512 | resolvedConfig.dockerImageBaseName=comfyui | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=large | recommendedInstance.id=gpu-a6000 | recommendedInstance.name=GPU High VRAM (A6000) | recommendedInstance.suitableFor=30B+ LLMs (quantized 70B), Video Gen, 48GB VRAM | recommendedInstance.vCPU=9 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA RTX A6000 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-high-vram-value | recommendedInstance.deployCost=0.49; seo=slug=qwen-image-2512 | modelType=image | h1=Deploy Qwen-Image-2512 | aboutBlurb=Qwen-Image-2512 is the #1 ranked open-source image generation model; It excels at realistic human faces, accurate text rendering, and complex compositions; Built by Alibaba's Tongyi Lab; | useCases=Marketing and advertising visuals, Product photography, Text-heavy graphic design, Realistic portrait generation | relatedGroups=flux-2-klein, z-image-turbo, flux | configKeys=qwen-image-2512; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-20 nearby=hidream-i1-dev-fp8; relation=same-type-neighbor; group=hidream; size=Dev FP8 (Recommended)/large; instance=gpu-a6000; hourly=0.49; monthly=352.8; delta-vs-flux2-klein-9b-fp8=0.1; delta-vs-flux2-klein-4b=0.1; template=comfyui-hidream; docker=comfyui; template-match-a=no; template-match-b=no; term-overlap-a=4; term-overlap-b=4; source=hidream-i1-dev-fp8; description=17B parameter model with excellent prompt following; Uses Flux architecture;; useCases=Prompt-accurate image generation; Creative illustration; Commercial image production; Fine-tuning base model; config=id=hidream-i1-dev-fp8 | name=HiDream I1 Dev FP8 | displayName=HiDream I1 | sizeVariant=Dev FP8 (Recommended) | templateName=comfyui-hidream | dockerImageBaseName=comfyui | dockerImageTag=hidream | modelSize=large | popularity=8 | modelGroup=hidream | groupPopularity=85 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-hidream | resolvedConfig.dockerImageBaseName=comfyui | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=large | recommendedInstance.id=gpu-a6000 | recommendedInstance.name=GPU High VRAM (A6000) | recommendedInstance.suitableFor=30B+ LLMs (quantized 70B), Video Gen, 48GB VRAM | recommendedInstance.vCPU=9 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA RTX A6000 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-high-vram-value | recommendedInstance.deployCost=0.49; seo=slug=hidream-i1 | modelType=image | h1=Deploy HiDream I1 | aboutBlurb=HiDream I1 is a 17B parameter image generation model with excellent prompt following; Available in Dev, Full, and Fast variants with FP8 quantization for efficient deployment on consumer GPUs; | useCases=Prompt-accurate image generation, Creative illustration, Commercial image production, Fine-tuning base model | relatedGroups=flux-2, qwen-image-2512, flux | configKeys=hidream-i1-dev-fp8, hidream-i1-full-fp8, hidream-i1-fast-fp8; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-21 nearby=qwen-image-edit; relation=same-type-neighbor; group=qwen-image; size=Edit 2511/large; instance=gpu-a6000; hourly=0.49; monthly=352.8; delta-vs-flux2-klein-9b-fp8=0.1; delta-vs-flux2-klein-4b=0.1; template=comfyui-qwen; docker=comfyui; template-match-a=no; template-match-b=no; term-overlap-a=3; term-overlap-b=3; source=qwen-image-edit; description=AI product staging and image editing; Upload a photo, describe the scene you want; 4-step Lightning;; useCases=Image editing and inpainting; Bilingual prompt support; Text-guided image modification; Chinese-language image generation; config=id=qwen-image-edit | name=Qwen Image Edit 2511 | displayName=Qwen Image | sizeVariant=Edit 2511 | templateName=comfyui-qwen | dockerImageBaseName=comfyui | dockerImageTag=qwen-image-gen | modelSize=large | popularity=8 | modelGroup=qwen-image | groupPopularity=80 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-qwen | resolvedConfig.dockerImageBaseName=comfyui | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=large | recommendedInstance.id=gpu-a6000 | recommendedInstance.name=GPU High VRAM (A6000) | recommendedInstance.suitableFor=30B+ LLMs (quantized 70B), Video Gen, 48GB VRAM | recommendedInstance.vCPU=9 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA RTX A6000 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-high-vram-value | recommendedInstance.deployCost=0.49; seo=slug=qwen-image | modelType=image | h1=Deploy Qwen Image | aboutBlurb=Qwen Image models from Alibaba Tongyi provide precise text and semantic image editing alongside high-quality generation; They support bilingual Chinese and English prompts for global workflows; | useCases=Image editing and inpainting, Bilingual prompt support, Text-guided image modification, Chinese-language image generation | relatedGroups=qwen-image-2512, flux, stable-diffusion | configKeys=qwen-image-edit, qwen-image-gen; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-22 nearby=flux-story-krea; relation=same-type-neighbor; group=story-diffusion; size=large/large; instance=gpu-a6000; hourly=0.49; monthly=352.8; delta-vs-flux2-klein-9b-fp8=0.1; delta-vs-flux2-klein-4b=0.1; template=comfyui-story; docker=comfyui-story; template-match-a=no; template-match-b=no; term-overlap-a=0; term-overlap-b=0; source=flux-story-krea; description=Character-consistent sequential image generation with natural Krea aesthetics; Perfect for visual storytelling, comics, and brand mascots;; useCases=; config=id=flux-story-krea | name=Flux.1 Krea + StoryDiffusion | templateName=comfyui-story | dockerImageBaseName=comfyui-story | modelSize=large | popularity=7 | modelGroup=story-diffusion | groupPopularity=75 | templateOnly=true | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-story | resolvedConfig.dockerImageBaseName=comfyui-story | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=large | recommendedInstance.id=gpu-a6000 | recommendedInstance.name=GPU High VRAM (A6000) | recommendedInstance.suitableFor=30B+ LLMs (quantized 70B), Video Gen, 48GB VRAM | recommendedInstance.vCPU=9 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA RTX A6000 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-high-vram-value | recommendedInstance.deployCost=0.49; seo=; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-23 nearby=hidream-i1-full-fp8; relation=same-type-neighbor; group=hidream; size=Full FP8/large; instance=gpu-a6000; hourly=0.49; monthly=352.8; delta-vs-flux2-klein-9b-fp8=0.1; delta-vs-flux2-klein-4b=0.1; template=comfyui-hidream; docker=comfyui; template-match-a=no; template-match-b=no; term-overlap-a=4; term-overlap-b=4; source=hidream-i1-full-fp8; description=Full quality 17B model with FP8 quantization;; useCases=Prompt-accurate image generation; Creative illustration; Commercial image production; Fine-tuning base model; config=id=hidream-i1-full-fp8 | name=HiDream I1 Full FP8 | displayName=HiDream I1 | sizeVariant=Full FP8 | templateName=comfyui-hidream | dockerImageBaseName=comfyui | dockerImageTag=hidream | modelSize=large | popularity=7 | modelGroup=hidream | groupPopularity=85 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-hidream | resolvedConfig.dockerImageBaseName=comfyui | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=large | recommendedInstance.id=gpu-a6000 | recommendedInstance.name=GPU High VRAM (A6000) | recommendedInstance.suitableFor=30B+ LLMs (quantized 70B), Video Gen, 48GB VRAM | recommendedInstance.vCPU=9 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA RTX A6000 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-high-vram-value | recommendedInstance.deployCost=0.49; seo=slug=hidream-i1 | modelType=image | h1=Deploy HiDream I1 | aboutBlurb=HiDream I1 is a 17B parameter image generation model with excellent prompt following; Available in Dev, Full, and Fast variants with FP8 quantization for efficient deployment on consumer GPUs; | useCases=Prompt-accurate image generation, Creative illustration, Commercial image production, Fine-tuning base model | relatedGroups=flux-2, qwen-image-2512, flux | configKeys=hidream-i1-dev-fp8, hidream-i1-full-fp8, hidream-i1-fast-fp8; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
- flux2-klein-9b-fp8__flux2-klein-4b catalog-rank-24 nearby=hidream-i1-fast-fp8; relation=same-type-neighbor; group=hidream; size=Fast FP8 (4-step)/large; instance=gpu-a6000; hourly=0.49; monthly=352.8; delta-vs-flux2-klein-9b-fp8=0.1; delta-vs-flux2-klein-4b=0.1; template=comfyui-hidream; docker=comfyui; template-match-a=no; template-match-b=no; term-overlap-a=4; term-overlap-b=4; source=hidream-i1-fast-fp8; description=Distilled 4-step version for fast generation;; useCases=Prompt-accurate image generation; Creative illustration; Commercial image production; Fine-tuning base model; config=id=hidream-i1-fast-fp8 | name=HiDream I1 Fast FP8 | displayName=HiDream I1 | sizeVariant=Fast FP8 (4-step) | templateName=comfyui-hidream | dockerImageBaseName=comfyui | dockerImageTag=hidream | modelSize=large | popularity=6 | modelGroup=hidream | groupPopularity=85 | defaultConfig.templateName=comfyui-flux | defaultConfig.dockerImageBaseName=comfyui-flux | defaultConfig.containerPort=8188 | defaultConfig.volumeMountPath=/data | defaultConfig.modelSize=medium | resolvedConfig.templateName=comfyui-hidream | resolvedConfig.dockerImageBaseName=comfyui | resolvedConfig.containerPort=8188 | resolvedConfig.volumeMountPath=/data | resolvedConfig.modelSize=large | recommendedInstance.id=gpu-a6000 | recommendedInstance.name=GPU High VRAM (A6000) | recommendedInstance.suitableFor=30B+ LLMs (quantized 70B), Video Gen, 48GB VRAM | recommendedInstance.vCPU=9 | recommendedInstance.ram=50 | recommendedInstance.disk=200 | recommendedInstance.gpuType=NVIDIA RTX A6000 | recommendedInstance.gpuCount=1 | recommendedInstance.type=gpu-high-vram-value | recommendedInstance.deployCost=0.49; seo=slug=hidream-i1 | modelType=image | h1=Deploy HiDream I1 | aboutBlurb=HiDream I1 is a 17B parameter image generation model with excellent prompt following; Available in Dev, Full, and Fast variants with FP8 quantization for efficient deployment on consumer GPUs; | useCases=Prompt-accurate image generation, Creative illustration, Commercial image production, Fine-tuning base model | relatedGroups=flux-2, qwen-image-2512, flux | configKeys=hidream-i1-dev-fp8, hidream-i1-full-fp8, hidream-i1-fast-fp8; files=src/config/modelConfig.js, src/config/modelSeoData.js, src/config/instanceConfig.js
FAQ
What GPU runs flux2-klein-9b-fp8 in the flux2-klein-9b-fp8 vs flux2-klein-4b comparison?
For flux2-klein-9b-fp8 vs flux2-klein-4b, flux2-klein-9b-fp8 uses GPU Efficient (L4) with 24GB VRAM at $0.39/hr.
What GPU runs flux2-klein-4b in the flux2-klein-9b-fp8 vs flux2-klein-4b comparison?
For flux2-klein-9b-fp8 vs flux2-klein-4b, flux2-klein-4b uses GPU Efficient (L4) with 24GB VRAM at $0.39/hr.
Which is cheaper to run 24/7 for flux2-klein-9b-fp8 vs flux2-klein-4b?
For flux2-klein-9b-fp8 vs flux2-klein-4b, flux2-klein-9b-fp8 is cheaper at $280.80/month vs $280.80/month.