gpurank
esc
  • Nothing named “”. Press Enter to search the offers instead.
↑↓ to move · ↵ to open · esc to close0 entries
Black Forest Labs · Image generation · 5.6B parameters · FLUX Kommunity License v1.0

FLUX 3 Image Klein

Needs 7 GB at FP8 / INT8. The smallest real configuration today is 1× RTX 3060 at Vast.ai, $0.04 per hour, about $26 a month at 730 hours.

Parameters5.6Bdense
CategoryImage generationsized on weights + headroom
Checkpoint43 GBBF16, as published
LicenseFLUX Kommunity License v1.0check before commercial use
ReleasedOct 15, 2026documented date
Smallest configuration · FP8 / INT8
1× RTX 306012 GB VRAM · 56% used
Per hour
$0.04
Per month
$26
Cheapest at
Vast.ai
All RTX 3060 offers →

What is documented about FLUX 3 Image Klein

news checked 2026-10-27

Step-distilled (16 steps, no guidance). Text-to-image, editing and multi-reference (up to 10 images), bounding-box layout, native 4K. Files: transformer 11.10 GB BF16 (HF metadata 5.55B params), text encoder Qwen3-VL-4B-Instruct 8.88 GB, FLUX.2 VAE 0.34 GB; fp8r variant 6.27 GB, nvfp4 5.28 GB. Card says it is not meant to run without a prompt upsampler (see FLUX 3 Image Klein Prompt Upsampler). Speed (BFL): INT8 1 MP in ComfyUI about 4 s on RTX 5090, under 9 s on RTX 4070; NVFP4 under 0.6 s on GB300. Also covers image editing (single checkpoint). HF repo created 2026-10-13, public launch 2026-10-15.

  • https://huggingface.co/black-forest-labs/flux-3-image-klein
  • https://huggingface.co/api/models/black-forest-labs/flux-3-image-klein?blobs=true
  • https://bfl.ai/blog/flux-3-image-klein
Memory · 3 precisions

How much VRAM it needs

Weights plus 20% runtime overhead. For image generation models, activations also depend on resolution, clip length and batch size; the study's per-GPU estimates below give the practical configuration.

PrecisionWeightsWeights + headroom
BF16published checkpoint43 GB52 GB
FP8 / INT8estimated6 GB7 GB
INT4 / MXFP4estimated3 GB5 GB
Cross-checked on Hugging Face Oct 28, 2026: 5.55B parameters in the safetensors headers, 43.3 GB of weight files. Method: weights as published at the native precision, verified on Hugging Face 2026-10-28; other precisions estimated at 1 (8-bit) and 0.6 (4-bit) bytes per parameter; × 1.2 runtime headroom; KV cache not added (architecture not readable).
Smallest configuration per GPU · 110 GPUs · cheapest eligible on-demand offer per GPU count, every provider on file

What it costs to run, GPU by GPU

The fewest GPUs of one type whose total memory fits the model at short context, priced from the cheapest eligible on-demand offer for exactly that count (45-day window, newest reading per provider), with the next providers beside it. Monthly = 730 hours.

GPUSmallest configVRAM used$ / hour$ / monthCheapest at
RTX 3060Best fit1× 12 GB12 GB total56%$0.04$26Vast.aiTrust 78 · read Nov 1 · receipt · also Clore.ai $0.05, gcube $0.05
RTX A20001× 12 GB12 GB total56%$0.04$26Vast.aiTrust 78 · read Nov 1 · receipt · also gcube $0.09
Tesla P41× 8 GB8 GB total84%$0.04$29Vast.aiTrust 78 · read Nov 1 · receipt · also Akash Network $0.06, gcube $0.07
T41× 16 GB16 GB total42%$0.04$29Vast.aiTrust 78 · read Nov 1 · receipt · also iwinv $0.20, Theta EdgeCloud $0.20
Tesla P401× 24 GB24 GB total28%$0.04$30Clore.aiTrust 46 · read Nov 1 · receipt · also gcube $0.07, Vast.ai $0.11
GTX Titan X1× 12 GB12 GB total56%$0.04$31Vast.aiTrust 78 · read Nov 1 · receipt
Titan Xp1× 12 GB12 GB total56%$0.04$31Vast.aiTrust 78 · read Nov 1 · receipt
GTX 10801× 8 GB8 GB total84%$0.04$32Vast.aiTrust 78 · read Nov 1 · receipt · also Edgevana $0.06
Tesla P1001× 16 GB16 GB total42%$0.04$33Vast.aiTrust 78 · read Nov 1 · receipt · also gcube $0.07
GTX 10701× 8 GB8 GB total84%$0.05$36Vast.aiTrust 78 · read Nov 1 · receipt · also Clore.ai $0.08
RTX 40601× 8 GB8 GB total84%$0.05$40Vast.aiTrust 78 · read Nov 1 · receipt · also Clore.ai $0.08, Theta EdgeCloud $0.09
RTX 20701× 8 GB8 GB total84%$0.06$41Vast.aiTrust 78 · read Nov 1 · receipt · also gcube $0.06, Clore.ai $0.21
RTX 3060 Ti1× 8 GB8 GB total84%$0.06$41Vast.aiTrust 78 · read Nov 1 · receipt · also gcube $0.06, Clore.ai $0.06
RTX 2060 Super1× 8 GB8 GB total84%$0.06$41gcubenot scored · read Nov 1 · receipt · also Clore.ai $0.06, Vast.ai $0.07
Quadro P20002× 5 GB10 GB total67%$0.06$41Vast.aiTrust 78 · read Nov 1 · receipt
GTX 1070 Ti1× 8 GB8 GB total84%$0.06$44Vast.aiTrust 78 · read Nov 1 · receipt · also Clore.ai $0.10
RTX 2070 Super1× 8 GB8 GB total84%$0.06$44gcubenot scored · read Nov 1 · receipt · also Vast.ai $0.06, Theta EdgeCloud $0.09
RTX 30701× 8 GB8 GB total84%$0.06$44gcubenot scored · read Nov 1 · receipt · also Clore.ai $0.07, Theta EdgeCloud $0.08
GTX 1080 Ti1× 11 GB11 GB total61%$0.06$45Vast.aiTrust 78 · read Nov 1 · receipt · also Edgevana $0.08
RTX A40001× 16 GB16 GB total42%$0.06$46Clore.aiTrust 46 · read Nov 1 · receipt · also Vast.ai $0.07, Edgevana $0.10
Tesla V100 16GB1× 16 GB16 GB total42%$0.06$46Clore.aiTrust 46 · read Nov 1 · receipt · also Vast.ai $0.08, CloudRift $0.28
RTX 20801× 8 GB8 GB total84%$0.06$47gcubenot scored · read Nov 1 · receipt · also Clore.ai $0.08, Vast.ai $0.09
RTX 2080 Super1× 8 GB8 GB total84%$0.06$47gcubenot scored · read Nov 1 · receipt · also Clore.ai $0.13
RTX 2080 Ti1× 11 GB11 GB total61%$0.07$49gcubenot scored · read Nov 1 · receipt · also Vast.ai $0.07, Edgevana $0.09
The 24 cheapest of 104 GPU types that fit. Every GPU, any precision and context, in the finder