{"slug": "ornith-1-0-self-scaffolding-llms-for-agentic-coding", "title": "Ornith-1.0: Self-Scaffolding LLMs for Agentic Coding", "summary": "DeepReinforce released Ornith-1.0, a family of self-scaffolding LLMs for agentic coding built on Gemma 4 and Qwen 3.5, achieving state-of-the-art performance among open-source models of comparable size on coding benchmarks. The models, ranging from 9B to 397B parameters, are available under Apache 2.0 licenses and demonstrate proficient agentic tool use in initial tests.", "body_md": "[Ornith-1.0: Self-Scaffolding LLMs for Agentic Coding](https://deep-reinforce.com/ornith_1_0.html)\n\n[...] with variants including 9B Dense, 31B Dense, 35B MoE, and 397B MoE. Built on top of pretrained Gemma 4 and Qwen 3.5, it achieves state-of-the-art performance among open-source models of comparable size on coding benchmarks.\n\nAs far as I can tell the licenses of those underlying models is compatible with being used in this way - Gemma 4 is Apache 2.0 licensed (and not bound by the janky additional [Gemma Terms of Use](https://ai.google.dev/gemma/terms) that afflicted the previous Gemma models) and Qwen 3.5 is Apache 2.0 licensed as well.\n\nI've been running the model using LM Studio and the [ornith-1.0-35b-Q4_K_M.gguf](https://huggingface.co/deepreinforce-ai/Ornith-1.0-35B-GGUF) (20GB) GGUF, hooked up to [Pi](https://pi.dev/). Initial impressions are very good - it seems to be able to run the agent harness over many tool calls in a proficient way.\n\nHere's [a terminal session](https://gisthost.github.io/?35da4d9ce7f0c27124c67655a0dc9e5d) where I asked it to \"find the code that decodes the actor cookie\" and then \"find the code that opens the insert dialog when thebutton is clicked\" against a Datasette checkout, which it handled with ease.\n\nI also had it [draw this pelican](https://gist.github.com/simonw/1869e1bbcafe5bcad0f26351f6a978a6), which came out at 103 tokens/second:\n\nIt's a little bit mangled but the pelican is clearly a pelican.\n\nI couldn't find much information about DeepReinforce themselves. The earliest paper I could find from the was [CUDA-L1: Improving CUDA Optimization via Contrastive Reinforcement Learning](https://arxiv.org/abs/2507.14111) from June 2025.\n\nTags: [ai](https://simonwillison.net/tags/ai), [generative-ai](https://simonwillison.net/tags/generative-ai), [local-llms](https://simonwillison.net/tags/local-llms), [llms](https://simonwillison.net/tags/llms), [qwen](https://simonwillison.net/tags/qwen), [pelican-riding-a-bicycle](https://simonwillison.net/tags/pelican-riding-a-bicycle), [gemma](https://simonwillison.net/tags/gemma), [llm-release](https://simonwillison.net/tags/llm-release), [lm-studio](https://simonwillison.net/tags/lm-studio)", "url": "https://wpnews.pro/news/ornith-1-0-self-scaffolding-llms-for-agentic-coding", "canonical_source": "https://simonwillison.net/2026/Jun/29/ornith/#atom-everything", "published_at": "2026-06-29 16:17:59+00:00", "updated_at": "2026-06-29 16:49:39.548483+00:00", "lang": "en", "topics": ["large-language-models", "ai-agents", "ai-research", "ai-tools", "developer-tools"], "entities": ["DeepReinforce", "Ornith-1.0", "Gemma 4", "Qwen 3.5", "LM Studio", "Pi", "Datasette", "Hugging Face"], "alternates": {"html": "https://wpnews.pro/news/ornith-1-0-self-scaffolding-llms-for-agentic-coding", "markdown": "https://wpnews.pro/news/ornith-1-0-self-scaffolding-llms-for-agentic-coding.md", "text": "https://wpnews.pro/news/ornith-1-0-self-scaffolding-llms-for-agentic-coding.txt", "jsonld": "https://wpnews.pro/news/ornith-1-0-self-scaffolding-llms-for-agentic-coding.jsonld"}}