{"slug": "ornith", "name": "Ornith", "category": "ai-models", "category_name": "AI models", "description": "", "recommenders": 14, "commenters": 22, "warners": 1, "score": 14.287, "rising": 7.36, "first": "2026-06-25", "latest": "2026-10-15", "why": "my new daily driver is Ornith-1.5-35B-A3B-MLX-4bit. I recommend giving that a whirl", "why_is_quote": true, "has_verdict": false, "sentiment": {"recommend": 14, "neutral": 7, "compare": 0, "warn": 1}, "spark": [0, 0, 0, 0, 0, 0, 0, 3, 2, 4, 2, 3], "listed": true, "loved": [], "caveat": "", "short_quote": "my new daily driver is Ornith-1.5-35B-A3B-MLX-4bit. I recommend giving that a whirl", "kind": "", "icon": false, "domain": "deep-reinforce.com", "github_repo": null, "official_url": "", "pricing": "", "aliases": ["Ornith-1.0", "Ornith", "Ornith-1.5", "Ornith-1.5-35B-A3B", "Ornith 1.5 35B", "Ornith-1.5-35B-MTP-UD-APEX-GGUF", "Ornith-1.5-35B-A3B-oQ4e-mtp", "Ornith1.5-35B-A3B", "Ornith 9B", "Ornith 1.0 35B", "Ornith-1.5-9B", "Ornith 9G", "ornith 1.5 35b", "Ornith1"], "verdict": null, "praise": [{"label": "", "summary": "", "quotes": [{"id": 801563970, "item_id": 50097748, "url": "https://news.ycombinator.com/item?id=50097748", "by": "otekengineering", "date": "2026-10-15", "version": "1.5", "stance": "recommend", "quote": "it's usable for real coding and sys admin work, and works great in opencode", "excerpt": "tried opencode for the first time to use the $200 in api credits that come with claude max but can't be used in claude code. it's great. the free nemotron 3 model is even decent, i love seeing agentic coding accessible to students and hobbiests and folks that can't drop $200/mo. i used the bonus anthropic credits to get ornith 1.5 35b q8 running on a jetson. it's no opus 5, but it's usable for real coding and sys adm", "context": "", "use_case": "local coding and sysadmin on Jetson", "switched_from": "", "story": "OpenCode 2.0"}, {"id": 799992370, "item_id": 49999523, "url": "https://news.ycombinator.com/item?id=49999523", "by": "pcf", "date": "2026-10-07", "version": "1.5", "stance": "recommend", "quote": "I run these two Qwen model variations successfully", "excerpt": "I have a MacBook Air M2 with 24 GB RAM and I run these two Qwen model variations successfully. Try them yourself: * https://huggingface.co/Youssofal/Ternary-Bonsai-2-27B-MTPLX-Optimized-Speed * https://huggingface.co/gbuzhf/Ornith-1.5-35B-MTP-UD-APEX-GGUF So far just testing them for fun in LM Studio. I'm not a coder, so I'm testing them more on logic and language tasks. But yes, I do have a Mac Studio M3 Ultra with ", "context": "", "use_case": "", "switched_from": "", "story": "Anthropic Subscriptions Offer 5x+ More Value Than OpenAI"}, {"id": 799992372, "item_id": 49999523, "url": "https://news.ycombinator.com/item?id=49999523", "by": "pcf", "date": "2026-10-07", "version": "1.5", "stance": "recommend", "quote": "where I run the less quantised versions of these for higher quality results", "excerpt": "I have a MacBook Air M2 with 24 GB RAM and I run these two Qwen model variations successfully. Try them yourself: * https://huggingface.co/Youssofal/Ternary-Bonsai-2-27B-MTPLX-Optimized-Speed * https://huggingface.co/gbuzhf/Ornith-1.5-35B-MTP-UD-APEX-GGUF So far just testing them for fun in LM Studio. I'm not a coder, so I'm testing them more on logic and language tasks. But yes, I do have a Mac Studio M3 Ultra with ", "context": "", "use_case": "", "switched_from": "", "story": "Anthropic Subscriptions Offer 5x+ More Value Than OpenAI"}, {"id": 798894002, "item_id": 49930875, "url": "https://news.ycombinator.com/item?id=49930875", "by": "Otterly99", "date": "2026-10-02", "version": "1.5", "stance": "recommend", "quote": "(my personnal favorite)", "excerpt": "With 8GB I would recommend quantized versions of 9B models such as these: - https://huggingface.co/unsloth/Qwen3.5-9B-GGUF - https://huggingface.co/empero-ai/Qwen3.8-9B-Distill-GGUF (unofficial Qwen 3.8-9B) - https://huggingface.co/ornith-ai/Ornith-1.5-9B-GGUF (my personnal favorite)", "context": "", "use_case": "", "switched_from": "", "story": "Pi 1.0"}, {"id": 794644656, "item_id": 49665291, "url": "https://news.ycombinator.com/item?id=49665291", "by": "nasutton12", "date": "2026-09-11", "version": null, "stance": "recommend", "quote": "i still miss the speed of that moe", "excerpt": "the early 1.x versions of chad were tied to ornith. i still miss the speed of that moe. https://huggingface.co/nathansutton/Ornith-1.0-35B-UD-Q2_K_XL-MLX", "context": "", "use_case": "", "switched_from": "", "story": "Nine coding harnesses vs. your laptop"}]}], "caveats": [{"label": "", "summary": "", "quotes": [{"id": 789962544, "item_id": 49372659, "url": "https://news.ycombinator.com/item?id=49372659", "by": "readyblue", "date": "2026-08-20", "version": "1.5", "stance": "warn", "quote": "Please check your download links, it get 400 Bad Request", "excerpt": "Please check your download links, it get 400 Bad Request on https://huggingface.co/ornith-ai/Ornith-1.5-397B-GGUF/tree/main only mmproj can be downloaded right now", "context": "", "use_case": "", "switched_from": "", "story": "Ornith-1.5: From Self-Scaffolding to Self-Improvement"}]}], "who": [{"by": "dexterlagan", "context": "runs on MBP M5 Max 48GB", "url": "https://news.ycombinator.com/item?id=49487694", "date": "2026-08-29"}, {"by": "pwython", "context": "64GB M4 Max user", "url": "https://news.ycombinator.com/item?id=49481526", "date": "2026-08-28"}, {"by": "JSR_FDED", "context": "runs on 16GB M1", "url": "https://news.ycombinator.com/item?id=49057204", "date": "2026-07-26"}, {"by": "Balinares", "context": "tests local models", "url": "https://news.ycombinator.com/item?id=48743104", "date": "2026-06-30"}], "tags": [], "use_cases": ["local coding and sysadmin on jetson", "non-coding tasks", "local model on macbook pro", "quick local inquiries", "local llm on mac", "local llm"], "compared": [{"slug": "qwen", "name": "Qwen", "co": 12, "switched_from": 3, "switched_to": 0}, {"slug": "gemma", "name": "Gemma", "co": 2, "switched_from": 0, "switched_to": 0}, {"slug": "opencode", "name": "OpenCode", "co": 2, "switched_from": 0, "switched_to": 0}, {"slug": "llama-cpp", "name": "llama.cpp", "co": 2, "switched_from": 0, "switched_to": 0}], "makers": [], "threads": [{"url": "https://news.ycombinator.com/item?id=48722052", "title": "Ornith-1.0: self-improving open-source models for agentic coding", "mentions": 4}, {"url": "https://news.ycombinator.com/item?id=49362401", "title": "Ornith-1.5: From Self-Scaffolding to Self-Improvement", "mentions": 3}, {"url": "https://news.ycombinator.com/item?id=49975345", "title": "Anthropic Subscriptions Offer 5x+ More Value Than OpenAI", "mentions": 2}, {"url": "https://news.ycombinator.com/item?id=49953495", "title": "Run Qwen 3.8 Flash Next (125B) on consumer hardware (RTX 4090) at 100T/s", "mentions": 2}, {"url": "https://news.ycombinator.com/item?id=49479951", "title": "Run Qwen3.8 27B locally: real numbers from my Mac Studio", "mentions": 2}, {"url": "https://news.ycombinator.com/item?id=48721903", "title": "Qwen 3.6 27B is the sweet spot for local development", "mentions": 2}], "sentiment_by_month": {"2025-11": {}, "2025-12": {}, "2026-01": {}, "2026-02": {}, "2026-03": {}, "2026-04": {}, "2026-05": {}, "2026-06": {"recommend": 3, "neutral": 4}, "2026-07": {"neutral": 1, "recommend": 2}, "2026-08": {"recommend": 4, "warn": 1, "neutral": 1}, "2026-09": {"recommend": 2}, "2026-10": {"recommend": 3, "neutral": 1}}, "versions": [{"v": "1.5", "people": 10, "rec": 7, "warn": 1, "other": 2, "first": "2026-08-20"}, {"v": "1", "people": 7, "rec": 3, "warn": 0, "other": 4, "first": "2026-06-29"}], "version_unnamed": 4, "family_ref": null, "runs_on": null, "lines": [], "quotes": [{"id": 801563970, "item_id": 50097748, "url": "https://news.ycombinator.com/item?id=50097748", "by": "otekengineering", "date": "2026-10-15", "version": "1.5", "stance": "recommend", "quote": "it's usable for real coding and sys admin work, and works great in opencode", "excerpt": "tried opencode for the first time to use the $200 in api credits that come with claude max but can't be used in claude code. it's great. the free nemotron 3 model is even decent, i love seeing agentic coding accessible to students and hobbiests and folks that can't drop $200/mo. i used the bonus anthropic credits to get ornith 1.5 35b q8 running on a jetson. it's no opus 5, but it's usable for real coding and sys adm", "context": "", "use_case": "local coding and sysadmin on Jetson", "switched_from": "", "story": "OpenCode 2.0"}, {"id": 799992370, "item_id": 49999523, "url": "https://news.ycombinator.com/item?id=49999523", "by": "pcf", "date": "2026-10-07", "version": "1.5", "stance": "recommend", "quote": "I run these two Qwen model variations successfully", "excerpt": "I have a MacBook Air M2 with 24 GB RAM and I run these two Qwen model variations successfully. Try them yourself: * https://huggingface.co/Youssofal/Ternary-Bonsai-2-27B-MTPLX-Optimized-Speed * https://huggingface.co/gbuzhf/Ornith-1.5-35B-MTP-UD-APEX-GGUF So far just testing them for fun in LM Studio. I'm not a coder, so I'm testing them more on logic and language tasks. But yes, I do have a Mac Studio M3 Ultra with ", "context": "", "use_case": "", "switched_from": "", "story": "Anthropic Subscriptions Offer 5x+ More Value Than OpenAI"}, {"id": 799992372, "item_id": 49999523, "url": "https://news.ycombinator.com/item?id=49999523", "by": "pcf", "date": "2026-10-07", "version": "1.5", "stance": "recommend", "quote": "where I run the less quantised versions of these for higher quality results", "excerpt": "I have a MacBook Air M2 with 24 GB RAM and I run these two Qwen model variations successfully. Try them yourself: * https://huggingface.co/Youssofal/Ternary-Bonsai-2-27B-MTPLX-Optimized-Speed * https://huggingface.co/gbuzhf/Ornith-1.5-35B-MTP-UD-APEX-GGUF So far just testing them for fun in LM Studio. I'm not a coder, so I'm testing them more on logic and language tasks. But yes, I do have a Mac Studio M3 Ultra with ", "context": "", "use_case": "", "switched_from": "", "story": "Anthropic Subscriptions Offer 5x+ More Value Than OpenAI"}, {"id": 799294385, "item_id": 49955899, "url": "https://news.ycombinator.com/item?id=49955899", "by": "roscas", "date": "2026-10-04", "version": null, "stance": "compare", "quote": "I prefer https://ornith.ai/ornith_1_5.html to Qwen 3.8 not only because it is much faster on my hardware", "excerpt": "I prefer https://ornith.ai/ornith_1_5.html to Qwen 3.8 not only because it is much faster on my hardware but better responses. But this Qwen 3.8 Flash next coder is amazing running with Strata.", "context": "", "use_case": "", "switched_from": "", "story": "Run Qwen 3.8 Flash Next (125B) on consumer hardware (RTX 4090) at 100T/s"}, {"id": 799264418, "item_id": 49954026, "url": "https://news.ycombinator.com/item?id=49954026", "by": "roscas", "date": "2026-10-04", "version": "1.5", "stance": "neutral", "quote": "Any other task, I still use Ornith 1.5 35B that throws 20t/sec", "excerpt": "Coder version with 30t/sec on a Ryzen 3600x with 48GB of RAM with a nvidia 3080. This is not a very fast desktop. Memory speed is around 2000mhz only. My SSD is some of the worst SSD I've seen and 3080 had its days of glory. I still have code, chromium, librewolf and many other programs running. I have video streams running while I also watch tv and many times youtube videos. I use it with the browser that has a grea", "context": "", "use_case": "non-coding tasks", "switched_from": "", "story": "Run Qwen 3.8 Flash Next (125B) on consumer hardware (RTX 4090) at 100T/s"}, {"id": 798894002, "item_id": 49930875, "url": "https://news.ycombinator.com/item?id=49930875", "by": "Otterly99", "date": "2026-10-02", "version": "1.5", "stance": "recommend", "quote": "(my personnal favorite)", "excerpt": "With 8GB I would recommend quantized versions of 9B models such as these: - https://huggingface.co/unsloth/Qwen3.5-9B-GGUF - https://huggingface.co/empero-ai/Qwen3.8-9B-Distill-GGUF (unofficial Qwen 3.8-9B) - https://huggingface.co/ornith-ai/Ornith-1.5-9B-GGUF (my personnal favorite)", "context": "", "use_case": "", "switched_from": "", "story": "Pi 1.0"}, {"id": 794644656, "item_id": 49665291, "url": "https://news.ycombinator.com/item?id=49665291", "by": "nasutton12", "date": "2026-09-11", "version": null, "stance": "recommend", "quote": "i still miss the speed of that moe", "excerpt": "the early 1.x versions of chad were tied to ornith. i still miss the speed of that moe. https://huggingface.co/nathansutton/Ornith-1.0-35B-UD-Q2_K_XL-MLX", "context": "", "use_case": "", "switched_from": "", "story": "Nine coding harnesses vs. your laptop"}, {"id": 792554464, "item_id": 49534654, "url": "https://news.ycombinator.com/item?id=49534654", "by": "apexalpha", "date": "2026-09-02", "version": "1.5", "stance": "recommend", "quote": "it feels like an improvement.", "excerpt": "I have the exact same stack (macbook pro in stead of mini) and swapped from qwen to: https://huggingface.co/ornith-ai/Ornith-1.5-35B-A3B Literally today, but it feels like an improvement.", "context": "", "use_case": "Local model on MacBook Pro", "switched_from": "Qwen", "story": "My local model setup on an M4 Pro Mac Mini"}, {"id": 792135600, "item_id": 49508475, "url": "https://news.ycombinator.com/item?id=49508475", "by": "RomulusHill", "date": "2026-08-31", "version": "1.5", "stance": "neutral", "quote": "my inference company has actually started offering Ornith1.5 today for 9B and 35B A3B", "excerpt": "Hi Gertlabs, my inference company has actually started offering Ornith1.5 today for 9B and 35B A3B. If you are still interested, send me a message on X and I'll help you get started! https://x.com/romulushill https://scalattice.com/models/ornith-1.5-9b/ https://scalattice.com/models/ornith-1.5-35b-a3b/ https://scalattice.com/developers/", "context": "", "use_case": "", "switched_from": "", "story": "Ornith-1.5: From Self-Scaffolding to Self-Improvement"}, {"id": 791803104, "item_id": 49487694, "url": "https://news.ycombinator.com/item?id=49487694", "by": "dexterlagan", "date": "2026-08-29", "version": "1.5", "stance": "recommend", "quote": "It's now my go-to model for quick inquiries.", "excerpt": "Ornith 1.5 35B MoE is very close in most benchmarks, but runs at 100 tk/s where Qwen3.8 27B runs at 30 tk/s (MBP M5 Max 48GB). It's now my go-to model for quick inquiries. Ornith 1.0 had problems with world knowledge, but 1.5 seems very solid on that front. I'll be running some agentic tests today. Ornith 1.5 35B is so fast that my CPU doesn't even have time to heat up, where Qwen3.8 would turn my laptop into a toast", "context": "runs on MBP M5 Max 48GB", "use_case": "quick local inquiries", "switched_from": "Qwen3.8 27B", "story": "Run Qwen3.8 27B locally: real numbers from my Mac Studio"}, {"id": 791704416, "item_id": 49481526, "url": "https://news.ycombinator.com/item?id=49481526", "by": "pwython", "date": "2026-08-28", "version": "1.5", "stance": "recommend", "quote": "my new daily driver is Ornith-1.5-35B-A3B-MLX-4bit. I recommend giving that a whirl", "excerpt": "Yea Qwen3.8 wasn't fun to use on my 64GB M4 Max either (better than these numbers though), so my new daily driver is Ornith-1.5-35B-A3B-MLX-4bit. I recommend giving that a whirl if you're on similar hardware, it's definitely better than Qwen3.6 35b-a3b which was my go-to before. https://huggingface.co/ornith-ai/Ornith-1.5-35B-A3B", "context": "64GB M4 Max user", "use_case": "local LLM on Mac", "switched_from": "Qwen3.6", "story": "Run Qwen3.8 27B locally: real numbers from my Mac Studio"}, {"id": 791200913, "item_id": 49450057, "url": "https://news.ycombinator.com/item?id=49450057", "by": "Phemist", "date": "2026-08-26", "version": "1.5", "stance": "recommend", "quote": "Ornith1.5-35B-A3B I've been running (and quite happy about).", "excerpt": "They are already working on it. https://huggingface.co/unsloth/Qwen3.8-Flash-Next-GGUF https://unsloth.ai/docs/models/qwen3.8-next > You will need at least 75 GB of RAM or unified memory to run the model. Its smallest 1-bit quantized version is larger than usual because of the model\u2019s architecture so 1-bit isn't really 1-bit at all. However, this also means the quantization is less aggressive, allowing the model to r", "context": "", "use_case": "Local LLM", "switched_from": "", "story": "Qwen3.8-Flash-Next"}, {"id": 789962544, "item_id": 49372659, "url": "https://news.ycombinator.com/item?id=49372659", "by": "readyblue", "date": "2026-08-20", "version": "1.5", "stance": "warn", "quote": "Please check your download links, it get 400 Bad Request", "excerpt": "Please check your download links, it get 400 Bad Request on https://huggingface.co/ornith-ai/Ornith-1.5-397B-GGUF/tree/main only mmproj can be downloaded right now", "context": "", "use_case": "", "switched_from": "", "story": "Ornith-1.5: From Self-Scaffolding to Self-Improvement"}, {"id": 789827456, "item_id": 49364216, "url": "https://news.ycombinator.com/item?id=49364216", "by": "prometheus1992", "date": "2026-08-19", "version": "1", "stance": "recommend", "quote": "Ornith1 (9B) was a really nice model.", "excerpt": "Can't wait to try this. Ornith1 (9B) was a really nice model. I have been running it locally using - https://github.com/deepanwadhwa/samosa-chat", "context": "", "use_case": "local model", "switched_from": "", "story": "Ornith-1.5: From Self-Scaffolding to Self-Improvement"}, {"id": 784915264, "item_id": 49057204, "url": "https://news.ycombinator.com/item?id=49057204", "by": "JSR_FDED", "date": "2026-07-26", "version": "9", "stance": "recommend", "quote": "the subtlety and nuance it showed in its response was very surprising to me", "excerpt": "I use to think small models needed more handholding but I think it\u2019s more a question of age. Newer models do fine with less handholding regardless of size. Yesterday for fun I tried Ornith 9G on my 16GB M1 with OpenCode. It took 3.5 hours do so what Kimi does in a minute (prompt: \u201canalyze this codebase front end and back end and explain the auth mechanism\u201d), but the subtlety and nuance it showed in its response was v", "context": "runs on 16GB M1", "use_case": "codebase analysis on local machine", "switched_from": "", "story": "Claude Code Cut Their System Prompt by 80%. Does That Work for Small Models Too?"}, {"id": 782584320, "item_id": 48911520, "url": "https://news.ycombinator.com/item?id=48911520", "by": "erelong", "date": "2026-07-14", "version": "9B", "stance": "recommend", "quote": "it did a little better than Qwen 9B", "excerpt": "I was trying Ornith 9B locally (it's up on Ollama) which claims: > Ornith-1.0-9B, which can be easily deployed on edge devices, matches or exceeds the performance of much larger models such as Gemma 4-31B and Qwen 3.6 35B. https://deep-reinforce.com/ornith_1_0.html Only tried it so much so far; it did a little better than Qwen 9B", "context": "", "use_case": "local edge LLM", "switched_from": "", "story": "Bonsai 27B (1-bit LLM): The First 27B-Class Model to Run on a Phone"}, {"id": 780655954, "item_id": 48790997, "url": "https://news.ycombinator.com/item?id=48790997", "by": "lhl", "date": "2026-07-04", "version": "1", "stance": "neutral", "quote": "local/ornith-1.0-35b", "excerpt": "You got me curious, so I made a little harness comparison to my model test suite: Model Adapter Suite Score Passed Tasks --------------------------------- ------------ -------------- ------ ------ ----- local/ornith-1.0-35b little_coder aider_polyglot 36.0% 81/225 225 local/ornith-1.0-35b pi_devstack aider_polyglot 39.6% 89/225 225 local/ornith-1.0-35b pi_vanilla aider_polyglot 32.0% 72/225 225 Little Code does a lit", "context": "", "use_case": "", "switched_from": "", "story": "Ornith-1.0: self-improving open-source models for agentic coding"}, {"id": 779889664, "item_id": 48743104, "url": "https://news.ycombinator.com/item?id=48743104", "by": "Balinares", "date": "2026-06-30", "version": "1", "stance": "neutral", "quote": "There's definitely some latent capability in there that arguably exceeds Qwen 3.6", "excerpt": "Ok, I gave the 35B MoE weights a shot at Q6. It failed my go-to code reasoning benchmark (local models usually do), but delivered the most advanced output I've seen so far for my (non-agentic) coding execution benchmark. It was also full of bugs, though more in the form of coding mistakes than the deep logic bugs I usually see in local models. The code was also a bit haphazard, and some of those bugs could have been ", "context": "tests local models", "use_case": "local coding benchmark", "switched_from": "", "story": "Ornith-1.0: Self-Scaffolding LLMs for Agentic Coding"}, {"id": 779828258, "item_id": 48739266, "url": "https://news.ycombinator.com/item?id=48739266", "by": "nacs", "date": "2026-06-30", "version": "1", "stance": "neutral", "quote": "https://huggingface.co/deepreinforce-ai/Ornith-1.0-9B/blob/main/config.json#L3", "excerpt": "Only the 31b is Gemma. All the rest - including 9B - are Qwen 3.5/3.6: https://huggingface.co/deepreinforce-ai/Ornith-1.0-9B/blob/main/config.json#L3", "context": "", "use_case": "", "switched_from": "", "story": "Ornith-1.0: self-improving open-source models for agentic coding"}, {"id": 779659472, "item_id": 48728717, "url": "https://news.ycombinator.com/item?id=48728717", "by": "Narew", "date": "2026-06-29", "version": "1", "stance": "recommend", "quote": "From what I personally tested Ornith-1.0 35B is slightly better than Qwen-3.6 35B.", "excerpt": "From what I personally tested Ornith-1.0 35B is slightly better than Qwen-3.6 35B. My tests are tasks that consist of adding/modify feature in a big C++ codebase. The part that I find interesting is that the model is way faster than Qwen3.6 35B. It seems Ornith produce a smaller chain of thought. On my test it can be 3 time faster to produce the answer. I use it via llamacpp and codex-cli.", "context": "", "use_case": "C++ codebase feature work", "switched_from": "", "story": "Ornith-1.0: self-improving open-source models for agentic coding"}, {"id": 779625136, "item_id": 48726571, "url": "https://news.ycombinator.com/item?id=48726571", "by": "SwellJoe", "date": "2026-06-29", "version": null, "stance": "neutral", "quote": "", "excerpt": "The Ornith folks say they're doing that, but haven't released the Gemma-based 31b yet (https://github.com/deepreinforce-ai/Ornith-1). But, also, the Qwen-based 35b MoE Ornith version performs worse than Qwen 3.6 and Qwen AgentWorld on my benchmarks (which are focused on finding security bugs, so not exactly the same as agentic coding, but closely related skills). That said, the reason they're able to release Ornith b", "context": "", "use_case": "", "switched_from": "", "story": "Qwen 3.6 27B is the sweet spot for local development"}, {"id": 779611280, "item_id": 48725705, "url": "https://news.ycombinator.com/item?id=48725705", "by": "S0y", "date": "2026-06-29", "version": "1", "stance": "neutral", "quote": "Sure.", "excerpt": "Sure. https://deep-reinforce.com/ornith_1_0.html >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. >Ornith-1.0 is a self-improving training framework. Instead of relying on human-designed harnesses to drive solution generation in RL, Ornith-1.0 learns to generate both solution rollouts and the task-specific harn", "context": "", "use_case": "", "switched_from": "", "story": "Ornith-1.0: self-improving open-source models for agentic coding"}, {"id": 779563376, "item_id": 48722711, "url": "https://news.ycombinator.com/item?id=48722711", "by": "simonw", "date": "2026-06-29", "version": "1", "stance": "recommend", "quote": "I'm pretty impressed with it", "excerpt": "I've been trying Ornith 1.0 35B, I'm pretty impressed with it: https://simonwillison.net/2026/Jun/29/ornith/", "context": "", "use_case": "", "switched_from": "", "story": "Qwen 3.6 27B is the sweet spot for local development"}, {"id": 778792640, "item_id": 48674540, "url": "https://news.ycombinator.com/item?id=48674540", "by": "kathyxiao", "date": "2026-06-25", "version": null, "stance": "recommend", "quote": "Ornith achieves SOTA performance among open-source models of comparable size", "excerpt": "Ornith is a family of open-source models specialized for agentic coding. Ornith achieves SOTA performance among open-source models of comparable size on a variety of coding benchmarks (Terminal-Bench 2.1, SWE, NL2Repo, OpenClaw, SWE Atlas, etc) details: https://deep-reinforce.com/ornith_1_0.html", "context": "", "use_case": "Agentic coding model", "switched_from": "", "story": "New agentic coding SOTA models"}], "latest_version": {"label": "Ornith 1.5", "people": 10, "rec": 7, "warn": 1, "other": 2}}