DeepSeek
Hangzhou-based lab known for strong open-weight models.
DeepSeek releases highly competitive open-weight models, including the DeepSeek-V series, widely deployed on EU and US inference providers.
deepseek.comModels from DeepSeek
| Model | Context | Zero data retention | Training | Region | Provider |
|---|---|---|---|---|---|
| 1M | Enterprise | No | Multi | Alibaba Cloud | |
| 1M | Enterprise | No | Multi | Azure | |
| 1M | Enterprise | No | Multi | BytePlus | |
| 1M | Enterprise | No | US | DeepInfra | |
| 1M | Enterprise | No | Multi | Fireworks | |
| 1M | Enterprise | No | EU | Nebius | |
| 1M | Enterprise | No | US | Novita | |
| 1M | Zero data retention | No | EU | TensorX | |
| 1M | Zero data retention | No | US | Geodd | |
| 1M | Enterprise | No | Multi | Alibaba Cloud | |
| 1M | Enterprise | No | Multi | BytePlus | |
| 1M | Enterprise | No | US | DeepInfra | |
| 1M | Enterprise | No | Multi | Fireworks | |
| 1M | Zero data retention | No | US | Geodd | |
| 1M | Enterprise | No | US | Morph | |
| 1M | Enterprise | No | US | Novita | |
| 1M | Zero data retention | No | EU | Sference | |
| 1M | Zero data retention | No | EU | TensorX | |
| 164K | Enterprise | No | US | DeepInfra | |
| 164K | Enterprise | No | US | Novita | |
| 164K | Zero data retention | No | EU | TensorX | |
| 131K | Enterprise | No | Multi | BytePlus | |
| 32K | Zero data retention | No | US | Infercom | |
| 164K | Enterprise | No | US | DeepInfra | |
| 131K | Enterprise | No | US | Novita | |
| 164K | Enterprise | No | US | DeepInfra | |
| 164K | Zero data retention | No | EU | TensorX | |
| 131K | Enterprise | No | US | Novita | |
| 128K | Zero data retention | No | US | Infercom | |
| 64K | Enterprise | No | US | Novita | |
| 164K | Enterprise | No | US | DeepInfra | |
| 164K | Enterprise | No | US | Novita | |
| 164K | Enterprise | No | US | DeepInfra | |
| 64K | Enterprise | No | US | Novita | |
| 64K | Enterprise | No | US | Novita | |
| 131K | Enterprise | No | US | DeepInfra | |
| 8K | Enterprise | No | US | Novita | |
| 33K | Enterprise | No | US | Novita | |
| 164K | Enterprise | No | US | Novita | |
| 64K | Enterprise | No | US | Novita | |
| 128K | Enterprise | No | US | Novita | |
| 164K | Enterprise | No | US | DeepInfra | |
| 164K | Enterprise | No | US | Novita | |
| 164K | Zero data retention | No | EU | TensorX | |
| 128K | Enterprise | No | US | Novita | |
| 64K | Enterprise | No | US | Novita | |
| 33K | Enterprise | No | US | DeepInfra | |
| 160K | Enterprise | No | US | Novita | |
| 128K | Enterprise | No | US | Novita | |
| 64K | Enterprise | No | US | Novita | |
| 64K | Enterprise | No | US | Novita |
No longer available
DeepSeek models we used to route but the gateway no longer offers. Their pages stay up for reference — contact us if you need one back.