DeepReinforce Launches Ornith-1.0: Self-Constructed LLM for Agentic Coding
Decision Brief
What changedDeepReinforce releases its first open-source model, Ornith-1.0, based on Gemma 4 and Qwen 3.5, achieving state-of-the-art results across coding benchmarks.
Why it mattersIts self-architecture enables efficient multi-tool calling, offering valuable insights for developers building AI agent toolchains.
Who should careOpen-source model users
Affected stackQwen
Builder actionEvaluate
Source confidenceMedium · Reliable media or first-hand reporting
DeepReinforce released Ornith-1.0 under MIT license in four variants: 9B Dense, 31B Dense, 35B MoE, and 397B MoE. Built on Gemma 4 and Qwen 3.5, it tops open-source models of similar size in coding benchmarks. Tested via LM Studio with the 35B Q4_K_M GGUF (20GB), the model smoothly handles agent tasks like finding code to decode an actor cookie, generating at 103 tokens/s.
Summary basis: official / RSS sourceUnless it says 'full article read', this summary is based only on publicly available content — it never pretends to have read restricted originals.
Sources
- Simon Willison：Blog
Hands-on notes on LLM tools, local models, and practical AI engineering.
- Simon Willison：Blog