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 |
|---|---|---|---|---|---|---|
| jina-embeddings-v5-omni-nano jinaai | Feature extraction | ~2.2 GB | €0.04/h RTX 3060 | 1B | 6.5K | CC BY-NC 4.0 |
| linkup-sparseup-embed-v1 Linkup-Platform | Feature extraction | ~0.4 GB | €0.04/h RTX 3060 | 0.2B | 6.5K | Apache 2.0 |
| embeddinggemma-2 unsloth | Feature extraction | ~1.7 GB | €0.04/h RTX 3060 | 0.7B | 6.3K | Apache 2.0 |
| Hydrogen-1.0-MXFP4 deep-principle | Text generation | ~447 GB | — | 384B | 6K | MIT |
| text-to-video-ms-1.7b ali-vilab | Text to video | — | — | 1.4B | 5.7K | CC BY-NC 4.0 |
| NitroFusion2 ChenDY | Text to image | — | — | 6.2B | 5.6K | Apache 2.0 |
| whisper-large-v3-turbo-persian givia | Automatic speech recognition | — | — | 0.8B | 5.6K | CC BY 4.0 |
| granite-embedding-30m-sparse ibm-granite | Feature extraction | ~0.1 GB | €0.04/h RTX 3060 | 0B | 5.5K | Apache 2.0 |
| maya2-native maya-research | Text to speech | — | — | 2.5B | 5.2K | Apache 2.0 |
| Qwen3Guard-Stream-0.6B Qwen | Feature extraction | ~1.3 GB | €0.04/h RTX 3060 | 0.6B | 4.9K | Apache 2.0 |
| Qwen3.5-4B-3D-Coder-MTP-Repacked prithivMLmods | Image text to text | ~10 GB | €0.13/h RTX 5070 | 4.7B | 4.9K | Apache 2.0 |
| indic-transcribe-core bodhan-ai | Automatic speech recognition | — | — | 1.2B | 4.9K | Other |
| Carbon-X HuggingFaceBio | Text generation | ~33 GB | €0.62/h RTX 6000 Ada | 14.8B | 4.7K | Apache 2.0 |
| Ming-3.0-asr-tiny inclusionAI | Automatic speech recognition | — | — | 8.9B | 4.7K | MIT |
| Spark-X2.6-1.7B XHToken | Text generation | ~3.8 GB | €0.04/h RTX 3060 | 1.7B | 4.6K | Apache 2.0 |
| Clover-35B-A3B-Preview Diffuspace | Text generation | ~77 GB | €2.09/h RTX PRO 6000 | 34.7B | 4.4K | Apache 2.0 |
| MOSS-TTS OpenMOSS-Team | Text to speech | — | — | 8.5B | 4.4K | Apache 2.0 |
| Audio8-TTS-Preview-0.6b Edge0 | Text to speech | — | — | 0.6B | 4.3K | Apache 2.0 |
| nemotron-asr-streaming-farsi mehdi-hf | Automatic speech recognition | — | — | 0.6B | 4K | Other |
| SenseNova-Loop-9B-Preview sensenova | Image text to text | ~22 GB | €0.17/h RTX 3090 | 9.7B | 3.9K | Apache 2.0 |
| hibou-b histai | Feature extraction | ~0.2 GB | €0.04/h RTX 3060 | 0.1B | 3.8K | Apache 2.0 |
| AnimeGen-I2V aidealab | Image to video | — | — | 14.3B | 3.7K | Apache 2.0 |
| Giga-Embeddings-instruct ai-sage | Feature extraction | ~7.7 GB | €0.04/h RTX 3060 | 3.4B | 3.7K | MIT |
| Mage-VL-Cowork microsoft | Image text to text | ~11 GB | €0.13/h RTX 5070 | 4.7B | 3.7K | Apache 2.0 |
| llama-nv-embed-reasoning-3b nvidia | Feature extraction | ~7.2 GB | €0.04/h RTX 3060 | 3.2B | 3.6K | CC BY 4.0 |