Ornith 1.0 35B

Ornith-1.0-35B is a text-generation model specialized for agentic coding tasks, built as a 35B-parameter Mixture-of-Experts model post-trained on top of Qwen 3.5 and Gemma 4. It is the lightweight member of the Ornith family, designed for efficient single-GPU deployment while delivering state-of-the-art performance among open-source models of comparable size on coding benchmarks.

Self-Improving Training Framework

Ornith-1.0 employs reinforcement learning to jointly optimize the scaffold that drives solution rollouts and the resulting solution itself. By learning to generate both the search trajectories and the final outputs, the model discovers better solution paths and produces higher-quality code.

Benchmark Performance

The following table compares Ornith-1.0-35B against Qwen 3.5-35B, Qwen 3.6-35B, Gemma 4-31B, and Qwen 3.5-397B on agentic coding benchmarks.

Benchmark	Ornith-1.0-35B	Qwen3.5-35B	Qwen3.6-35B	Gemma4-31B	Qwen3.5-397B
Terminal-Bench 2.1 (Terminus-2)	64.2	41.4	52.5	42.1	53.5
Terminal-Bench 2.1 (Claude Code)	62.8	38.9	49.2	-	48.6
SWE-bench Verified	75.6	70	73.4	52	76.4
SWE-bench Pro	50.4	44.6	49.5	35.7	51.6
SWE-bench Multilingual	69.3	60.3	67.2	51.7	69.3
NL2Repo	34.6	20.5	29.4	15.5	36.8
Claw-eval Avg	69.8	65.4	68.7	48.5	70.7
SWE Atlas - QnA	37.1	13.2	15.5	-	20.4
SWE Atlas - RF	29.7	10.2	11.4	-	18.4
SWE Atlas - TW	27.8	9.8	13.3	-	18.5

Ornith-1.0-35B outperforms all comparable-size models on every listed benchmark and is competitive with the much larger Qwen 3.5-397B on SWE-bench Verified, SWE-bench Multilingual, and Claw-eval Avg.

Ornith-1.0-35B is a reasoning model: by default the assistant turn opens with a <think> … </think> block before the final answer. The model is released under the MIT license.