Another fantastic open source release.

Rohan Paul Twitter · Rohan Paul (@rohanpaul_ai) · 2026-06-25

DeepReinforce releases Ornith-1.0, an MIT-licensed family of open-source agentic coding LLMs, with the flagship 397B MoE model scoring 82.4 on SWE-Bench Verified.

Open original ↗

Appears in

Rapid AI Benchmark Improvement: Small Models and New Entrants Closing Capability Gaps

Extraction

Topics: open-source-llmagentic-codingmodel-releasecoding-benchmarks

Claims

DeepReinforce released Ornith-1.0, an MIT-licensed open-source family of agentic coding language models.
The flagship model, Ornith-1.0-397B, uses a Mixture-of-Experts architecture with 17B active parameters.
Ornith-1.0-397B scores 82.4 on SWE-Bench Verified, described as the most powerful model in the release.

Key quotes

DeepReinforce just dropped Ornith-1.0, an MIT-licensed open-source family of agentic coding LLMs.

The flagship Ornith-1.0-397B MoE (17B-active) is the most powerful model in the release, reporting 82.4 on SWE-Bench Verified.