Ornith 1.5 35B A3BDeepReinforce is an AI research startup founded by Jiwei Li. It is known for the CUDA-L1 and CUDA-L2 reinforcement learning work on GPU code optimisation and for the open-source Ornith family of agentic coding models, released under the MIT license.
- Type
- Private company
- Founder
- Jiwei Li
- AI lab
Model family
Ornith
DeepReinforce is an AI research startup founded by Jiwei Li. It is known for the CUDA-L1 and CUDA-L2 reinforcement learning work on GPU code optimisation and for the open-source Ornith family of agentic coding models, released under the MIT license.
Latest Ornith models
Models in this family
5 catalogued models
5 models · 1–5
| Select models to compare | Model | Released | Context | Input /M | Output /M | Parameters | Type | I/O | Openness |
|---|---|---|---|---|---|---|---|---|---|
| Ornith 1.5 35B A3BDeepReinforce | August 18, 2026 | 262K | — | — | 35B | Text | text, image → text | Open weights | |
| Ornith 1.0 31BDeepReinforce | June 25, 2026 | 262K | — | — | 31B | Text | text, image → text | Open weights | |
| Ornith 1.0 35BDeepReinforce | June 25, 2026 | 262K | — | — | 35B | Text | text, image → text | Open weights | |
| Ornith 1.0 397BDeepReinforce | June 25, 2026 | 262K | — | — | 397B | Text | text, image → text | Open weights | |
| Ornith 1.0 9BDeepReinforce | June 25, 2026 | 262K | — | — | 9B | Text | text, image → text | Open weights |