Fibo 2.0 Dev
Needs 10 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.
- Per hour
- $0.04
- Per month
- $26
- Cheapest at
Vast.ai
What is documented about Fibo 2.0 Dev
news checked 2026-10-278B single-stream flow-matching DiT (44 blocks) for text-to-image and editing in one set of weights, driven by long structured JSON captions; trained only on licensed data. Files: transformer 16.26 GB BF16, text encoder Qwen3-VL-4B-Instruct 8.88 GB, FLUX.2 VAE 0.34 GB. 50 steps, guidance 5, about 1 MP. Card claims #1 open-weights image editing and #2 text-to-image on Artificial Analysis (as of 2026-10-20). HF repo created 2026-10-18; launch 2026-10-20.
- https://huggingface.co/briaai/Fibo-2.0-dev
- https://huggingface.co/api/models/briaai/Fibo-2.0-dev?blobs=true
- https://comfyui-wiki.com/en/news/2026-10-20-fibo-2-0
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 per-GPU estimates below give the practical configuration.
| Precision | Weights | Weights + headroom |
|---|---|---|
| BF16published checkpoint | 26 GB | 31 GB |
| FP8 / INT8estimated | 8 GB | 10 GB |
| INT4 / MXFP4estimated | 5 GB | 6 GB |
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.
| GPU | Smallest config | VRAM used | $ / hour | $ / month | Cheapest at |
|---|---|---|---|---|---|
| RTX 3060Best fit | 1× 12 GB12 GB total | 81% | $0.04 | $26 | |
| RTX A2000 | 1× 12 GB12 GB total | 81% | $0.04 | $26 | |
| T4 | 1× 16 GB16 GB total | 61% | $0.04 | $29 | |
| Tesla P40 | 1× 24 GB24 GB total | 41% | $0.04 | $30 | |
| GTX Titan X | 1× 12 GB12 GB total | 81% | $0.04 | $31 | |
| Titan Xp | 1× 12 GB12 GB total | 81% | $0.04 | $31 | |
| Tesla P100 | 1× 16 GB16 GB total | 61% | $0.04 | $33 | |
| Quadro P2000 | 2× 5 GB10 GB total | 98% tight | $0.06 | $41 | |
| Tesla V100 16GB | 1× 16 GB16 GB total | 61% | $0.06 | $46 | |
| GTX 1080 Ti | 1× 11 GB11 GB total | 89% | $0.06 | $47 | |
| RTX 2080 Ti | 1× 11 GB11 GB total | 89% | $0.07 | $49 | |
| Tesla V100 32GB | 1× 32 GB32 GB total | 31% | $0.07 | $50 | |
| RTX PRO 2000 | 1× 16 GB16 GB total | 61% | $0.07 | $51 | |
| RTX A4000 | 1× 16 GB16 GB total | 61% | $0.07 | $53 | |
| Tesla M40 | 1× 24 GB24 GB total | 41% | $0.07 | $54 | |
| RTX 3080 | 1× 10 GB10 GB total | 98% tight | $0.07 | $55 | |
| RTX 3060 Ti | 2× 8 GB16 GB total | 61% | $0.08 | $61 | |
| RTX 5060 Ti | 1× 16 GB16 GB total | 61% | $0.08 | $61 | |
| RTX 4070 | 1× 12 GB12 GB total | 81% | $0.08 | $61 | |
| GTX 1660 | 2× 6 GB12 GB total | 81% | $0.09 | $64 | |
| RTX 3090 | 1× 24 GB24 GB total | 41% | $0.09 | $66 | |
| RTX 4070 Super | 1× 12 GB12 GB total | 81% | $0.10 | $70 | |
| GTX 1070 | 2× 8 GB16 GB total | 61% | $0.10 | $70 | |
| GTX 1080 | 2× 8 GB16 GB total | 61% | $0.10 | $76 | |
| The 24 cheapest of 100 GPU types that fit. Every GPU, any precision and context, in the finder | |||||
