BREAKING
DeepReinforce Releases Ornith-1.0
Self-Scaffolding Training
Conventional
fixed scaffold
●
Human-designed workflow
●
Optimizes solution only
Ornith-1.0
self-scaffolding
●
Drafts its own plan
●
Self-grades strategy and output
Ornith-9B Beats Bigger Models
Ornith-9B
43.1
Gemma4-31B
42.1
Ornith-35B
64.2
Ornith-397B
77.5
0
SWE-Bench Verified
0
context tokens
0
x
model variants
Run Locally via Ollama
1
Download from Hugging Face
↓
2
ollama run ornith:9b
↓
3
Looping under Ollama
↓
4
Prefer llama.cpp or vLLM
Real Projects Are the Real Test
AI NEWS BLITZ
DeepReinforce has launched Ornith-1.0, an open-source agentic coding model that runs locally.