Back to AIBriefs
LaunchAI ModelsDevelopers

DeepReinforce releases Ornith-1.0: coding model that learns its own RL scaffolds

DeepReinforce releases Ornith-1.0: coding model that learns its own RL scaffoldsHugging Face Models

Ornith-1.0 spans 9B to 397B parameters, available under MIT license. It is post-trained on Gemma 4 and Qwen 3.5, achieving state-of-the-art coding performance among open-source models of its size. The 397B MoE flagship matches frontier models on benchmarks.

deepreinforce-ai/Ornith-1.0-9B2 days agodeepreinforce-ai
deepreinforce-ai/Ornith-1.0-35B-GGUF2 days agodeepreinforce-ai
deepreinforce-ai/Ornith-1.0-35B2 days agodeepreinforce-ai
deepreinforce-ai/Ornith-1.0-397B2 days agodeepreinforce-ai
A US open source coding model just dropped and it's matching GLM on benchmarks. It's called...1 day agoVaibhav Sisinty
I'm blown away. Ornith-1.0 35b is the best local model I've ever run that doesn't require 200GB+...1 day agoAlex Finn
DeepReinforce has released Ornith-1.0, their new self-improving family of open-source models...2 days agoTestingCatalog News
DeepReinforce Releases Ornith-1.0: An Open-Source Coding Model Family That Learns Its Own RL Scaffolds2 days agoAsif Razzaq
Ornith-1.0 released on Hugging Face2 days agopaf1138Discuss
Stop waiting for Qwen3.7 Openweights.2 days ago-OpenSourcerDiscuss
Ornith 35B is great so far9 hours agoanubhav_200Discuss
·
Jun 25, 4:56 AM
DeepReinforce releases Ornith-1.0: coding model that learns its own RL scaffolds — AIBriefs