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 |
|---|---|---|---|---|---|---|
| Wan2.1-I2V-14B-720P Wan-AI | Image to video | — | — | 16.4B | 73.5K | Apache 2.0 |
| stable-diffusion-3.5-medium stabilityai | Text to image | — | — | 2.5B | 71.9K | Other |
| TurboWan2.1-T2V-1.3B-Diffusers IPostYellow | Text to video | — | — | 1.4B | 70.3K | Apache 2.0 |
| MOSS-Image-SFT-4-steps akhaliq | Text to image | — | — | 5.1B | 68.4K | Apache 2.0 |
| ideogram-4-fp8 ideogram-ai | Text to image | — | — | 9.3B | 67.5K | Other |
| clef Cloudflare | Image text to text | ~61 GB | €2.09/h RTX PRO 6000 | 27.4B | 66.3K | Apache 2.0 |
| dreamshaper-8 Lykon | Text to image | — | — | 0.9B | 65.3K | creativeml-openrail-m |
| AnimateLCM wangfuyun | Text to video | — | — | 0.5B | 63.3K | — |
| lingbot-world-fast robbyant | Image to video | — | — | 18.5B | 62.1K | Apache 2.0 |
| stable-diffusion-3-medium-diffusers stabilityai | Text to image | — | — | 2.1B | 62.1K | Other |
| orpheus-3b-0.1-ft-bf16 mlx-community | Text to speech | — | — | 3.3B | 61K | Apache 2.0 |
| Kromoss lodestones | Text to image | — | — | 5.1B | 57.1K | Apache 2.0 |
| granite-speech-5.0-470m-turboctc ibm-granite | Automatic speech recognition | — | — | 0.5B | 56.9K | Apache 2.0 |
| parakeet-ultra moondream | Automatic speech recognition | — | — | 0.6B | 53.8K | CC BY 4.0 |
| MiniMax-M3.1-Flash MiniMaxAI | Image text to text | ~250 GB | — | 220B | 53.1K | Other |
| moonshine-streaming-medium moonshine-ai | Automatic speech recognition | — | — | 0.3B | 50K | MIT |
| stable-diffusion-3.5-large stabilityai | Text to image | — | — | 8.1B | 49.2K | Other |
| Wan2.2-S2V-14B Wan-AI | Image to video | — | — | 16.3B | 49K | Apache 2.0 |
| s2-pro fishaudio | Text to speech | — | — | 4.6B | 47.7K | Other |
| Qwen-Image-2512 Qwen | Text to image | — | — | 20.4B | 47.4K | Apache 2.0 |
| mms-tts-eng facebook | Text to speech | — | — | 0B | 46.8K | CC BY-NC 4.0 |
| Confucius4-R2T2 netease-youdao | Automatic speech recognition | — | — | 2B | 46.6K | Other |
| Qwen3.8-Flash-Next-E384-FP4-V2 autotrust | Image text to text | ~178 GB | — | 155B | 46K | Other |
| flint RecursiveSI | Text generation | ~89 GB | €2.09/h RTX PRO 6000 | 39.8B | 46K | Apache 2.0 |
| Qwen-Image-2.1-Turbo Qwen | Text to image | — | — | 7.1B | 45.1K | Other |