DeepReinforce
Ornith 1.0 397B
Large coding-reasoning model for agentic software tasks and RL search
deepreinforce/ornith-1.0-397b
Context
262.1K
Output
—
Released
2026-06-25
Knowledge cutoff
—
Types
InputTextImage
OutputText
Reasoning
Supported
Tool use
Supported
Structured output
—
Attachments
Supported
Open weights
Supported
Provider pricing
No providers currently list this model.
Benchmarks
SWE-Bench Verified
82.4 · percent resolved
SWE-Bench Pro
62.2 · percent resolved
SWE-Bench Multilingual
78.9 · percent resolved
Terminal-Bench 2.1
77.5 · percent
Terminal-Bench 2.1
78.2 · percent
NL2Repo
48.2 · percent
Claw-eval
77.1 · percent