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 |
|---|---|---|---|---|---|---|
| VieNeu-TTS-v3-Turbo pnnbao-ump | Text to speech | — | — | 0.1B | 2.1M | Apache 2.0 |
| bge-reranker-large BAAI | Feature extraction | ~1.3 GB | €0.04/h RTX 3060 | 0.6B | 2M | MIT |
| Qwen2.5-Coder-14B-Instruct Qwen | Text generation | ~33 GB | €0.62/h RTX 6000 Ada | 14.8B | 1.9M | Apache 2.0 |
| NVIDIA-Nemotron-3-Nano-4B-BF16 nvidia | Text generation | ~8.9 GB | €0.09/h RTX 3080 | 4B | 1.9M | Other |
| wav2vec2-base-960h facebook | Automatic speech recognition | — | — | 0.1B | 1.8M | Apache 2.0 |
| SmolVLM2-500M-Video-Instruct HuggingFaceTB | Image text to text | ~1.1 GB | €0.04/h RTX 3060 | 0.5B | 1.8M | Apache 2.0 |
| Qwen3-14B-AWQ Qwen | Text generation | ~33 GB | €0.62/h RTX 6000 Ada | 14.8B | 1.8M | Apache 2.0 |
| DeepSeek-V4.1-Flash deepseek-ai | Image text to text | ~855 GB | — | 763B | 1.8M | MIT |
| multilingual-e5-large-instruct intfloat | Feature extraction | ~1.3 GB | €0.04/h RTX 3060 | 0.6B | 1.7M | MIT |
| Inkling thinkingmachines | Image text to text | ~2129 GB | — | 952B | 1.7M | Apache 2.0 |
| Gemma-4-26B-A4B-NVFP4 nvidia | Text generation | ~19 GB | €0.17/h RTX 3090 | 14.4B | 1.7M | Apache 2.0 |
| Qwen2.5-VL-3B-Instruct Qwen | Image text to text | ~8.4 GB | €0.09/h RTX 3080 | 3.8B | 1.7M | — |
| Qwen3-ASR-1.7B Qwen | Automatic speech recognition | — | — | 2.3B | 1.7M | Apache 2.0 |
| Qwen3.5-27B Qwen | Image text to text | ~62 GB | €2.09/h RTX PRO 6000 | 27.8B | 1.7M | Apache 2.0 |
| Qwen2.5-7B-Instruct-AWQ Qwen | Text generation | ~17 GB | €0.17/h RTX 3090 | 7.6B | 1.6M | Apache 2.0 |
| Qwen3-Coder-30B-A3B-Instruct Qwen | Text generation | ~68 GB | €2.09/h RTX PRO 6000 | 30.5B | 1.6M | Apache 2.0 |
| Qwen3-8B-AWQ Qwen | Text generation | ~18 GB | €0.17/h RTX 3090 | 8.2B | 1.6M | Apache 2.0 |
| DeepSeek-R1 deepseek-ai | Text generation | ~769 GB | — | 684B | 1.6M | MIT |
| Qwen3.5-35B-A3B Qwen | Image text to text | ~80 GB | €2.09/h RTX PRO 6000 | 36B | 1.6M | Apache 2.0 |
| NVIDIA-Nemotron-3-Nano-30B-A3B-BF16 nvidia | Text generation | ~71 GB | €2.09/h RTX PRO 6000 | 31.6B | 1.6M | Other |
| kokoro-inno-clone-tuner remsky | Text to speech | — | — | 0B | 1.6M | Apache 2.0 |
| JEV-27B-VL autotrust | Image text to text | ~62 GB | €2.09/h RTX PRO 6000 | 27.8B | 1.6M | Apache 2.0 |
| Qwen3-VL-8B-Instruct-FP8 Qwen | Image text to text | ~12 GB | €0.13/h RTX 5070 | 8.8B | 1.5M | Apache 2.0 |
| DeepSeek-V3 deepseek-ai | Text generation | ~769 GB | — | 684B | 1.5M | — |
| Qwen3-4B-Instruct-2507-FP8 Qwen | Text generation | ~5.8 GB | €0.04/h RTX 3060 | 4.4B | 1.5M | Apache 2.0 |