What does your model need?
Find an open model, see the GPU memory it takes to run or fine-tune, and the cheapest GPU that fits.
652 models. Memory marked ~ is an estimate from the model's size. Prices are full price for one GPU in stock, without VAT.
| Model | Task | GPU memory | Cheapest GPU that fits | Parameters | Downloads | License |
|---|---|---|---|---|---|---|
| LFM3-VL-2B LiquidAI | Image text to text | ~4.3 GB | €0.04/h RTX 3060 | 1.9B | 16.7K | Other |
| SheetSage2 m-a-p | Feature extraction | ~0.1 GB | €0.04/h RTX 3060 | 0.1B | 16.6K | CC BY-NC 4.0 |
| DeepSeek-V4.1-Flash-1017-UNCENSORED-FP8 dealignai | Image text to text | ~855 GB | — | 763B | 16.2K | MIT |
| Audio8-TTS-Realistic-Air Edge0 | Text to speech | — | — | 0.6B | 16.2K | Apache 2.0 |
| flux2-klein-9b-uncensored-text-encoder ponpoke | Text to image | — | — | 8.2B | 15.8K | Other |
| jina-embeddings-v2-base-zh jinaai | Feature extraction | ~0.4 GB | €0.04/h RTX 3060 | 0.2B | 15.7K | Apache 2.0 |
| phonon-40m-es gradium | Text to speech | — | — | 0.1B | 15.5K | CC BY-NC 4.0 |
| WeVisOne-VL-4B tencent | Image text to text | ~11 GB | €0.13/h RTX 5070 | 4.8B | 15.4K | Apache 2.0 |
| DreamX-World-5B GD-ML | Image to video | — | — | 5.3B | 15.1K | MIT |
| UniView Drexubery | Image to video | — | — | 17.8B | 14.9K | Apache 2.0 |
| C-RADIOv4-SO400M nvidia | Feature extraction | ~1 GB | €0.04/h RTX 3060 | 0.4B | 14.8K | Other |
| CogVideoX-2b zai-org | Text to video | — | — | 1.7B | 14.7K | Apache 2.0 |
| MERT-v2-FullSong m-a-p | Feature extraction | ~1.4 GB | €0.04/h RTX 3060 | 0.6B | 14.4K | CC BY-NC 4.0 |
| One-to-All-14b MochunniaN1 | Image to video | — | — | 18.8B | 14.1K | Apache 2.0 |
| stable-video-diffusion-img2vid-xt-1-1 vdo | Image to video | — | — | 1.5B | 13.9K | Other |
| WeVisOne-VL-2B tencent | Image text to text | ~5.5 GB | €0.04/h RTX 3060 | 2.4B | 13.7K | Apache 2.0 |
| NVIDIA-Nemotron-3.5-Transcribe-30B-A4B-NVFP4 nvidia | Automatic speech recognition | — | — | 18.7B | 13.7K | Other |
| Ming-3.0-tts-0.5B inclusionAI | Text to speech | — | — | 1.5B | 13.3K | MIT |
| NVIDIA-Nemotron-Labs-Dory1.0-4B-BF16 nvidia | Text generation | ~8.9 GB | €0.09/h RTX 3080 | 4B | 13.1K | Other |
| MOSS-Image-Base OpenMOSS-Team | Text to image | — | — | 5.1B | 12.6K | Apache 2.0 |
| Beam-1.0-501B-A23B reflection | Text generation | ~1120 GB | — | 501B | 12.6K | Apache 2.0 |
| phonon-40m-en gradium | Text to speech | — | — | 0.1B | 12.1K | CC BY-NC 4.0 |
| LiveAct Soul-AILab | Image to video | — | — | 18.9B | 11.9K | Apache 2.0 |
| DeepSeek-V4-Pro-Qwen3.5-9B-UNCENSORED-heretic xtools-at | Text generation | ~22 GB | €0.17/h RTX 3090 | 9.7B | 11.8K | Apache 2.0 |
| Hy4-preview tencent | Text generation | ~1743 GB | — | 780B | 11.5K | Apache 2.0 |