Ornith 1.0 35B
deepreinforce-ai/Ornith-1.0-35B-GGUF
published Jun 2026 · updated Jun 2026
Ornith 1.0 35B is a text-generation model for agentic coding, fine-tuned from Qwen 3.5 using a self-improving reinforcement learning framework.
specs
| Task | Text Generation (Agentic Coding) |
| Architecture | Mixture of Experts (MoE) |
| Parameters | 35B |
| License | MIT |
best for
- ·Automated software engineering and bug fixing via SWE-Bench tasks
- ·Terminal-based coding agents for multi-step development workflows
- ·Natural language to repository generation (NL2Repo)
FAQ
It is designed for agentic coding tasks such as automated bug fixing, repository generation, and terminal-based software engineering workflows.
It is a 35B parameter Mixture of Experts (MoE) model, post-trained on top of Qwen 3.5.
It is released under the MIT license, globally accessible with no regional limitations.
Use the gigarouter OpenAI-compatible endpoint with your API key, sending a chat completion request with the model name Ornith-1.0-35B.
It uses a chat template with system, user, and assistant messages, supporting tool calls via <tool_call> XML tags and reasoning blocks with <think> tags.
We're benchmarking and onboarding Ornith 1.0 35B as a hosted, OpenAI-compatible API. Sign in for free credit and be ready when it lands, or tell us you want it and we'll prioritize it.