{"v":1,"id":"model:hf:ornith-ai/Ornith-1.5-35B-A3B-GGUF","slug":"model-ornith-ai-ornith-1-5-35b-a3b-gguf","kind":"model","category":"llm","title":"Ornith-1.5-35B-A3B-GGUF","summary":"Chirp Chirp! 🐦 We are introducing Ornith-1.5, a major step toward building foundation models through end-to-end self-improvement.","source":{"provider":"hf","ref":"ornith-ai/Ornith-1.5-35B-A3B-GGUF","url":"https://huggingface.co/ornith-ai/Ornith-1.5-35B-A3B-GGUF","rev":"12393612fd4f730ff5aadc23e9b8f9648aa49ceb","fetchedAt":"2026-10-02T20:51:55.833Z","etag":"W/\"2c7b-rSKgDh2Xr28rX+3uKwjBRFEil44\""},"author":{"name":"ornith-ai","url":"https://huggingface.co/ornith-ai"},"license":{"spdx":"mit","raw":"mit","url":"https://huggingface.co/ornith-ai/Ornith-1.5-35B-A3B/blob/main/LICENSE","open":true},"metrics":{"downloads":3570548,"downloadsWeek":327838,"likes":473,"stars":130156,"openIssues":2528,"lastRelease":{"tag":"v0.5.0","at":"2026-09-23T20:50:06Z"},"pushedAt":"2026-10-02T20:17:08Z","takenAt":"2026-10-02T20:51:55.833Z"},"tags":["transformers","gguf","text-generation","endpoints_compatible","conversational"],"pipeline":"text-generation","links":{"github":"ggml-org/llama.cpp","npm":"node-llama-cpp"},"updatedAt":"2026-08-24T03:43:47.000Z","collectedAt":"2026-10-02T20:51:55.833Z","review":{"numbers":["3,570,548 downloads on Hugging Face","473 likes","license mit","20 GB for Ornith-1.5-35B-Q4_K_M.gguf","130,156 stars on ggml-org/llama.cpp","2,528 open issues and PRs","last release v0.5.0 on 2026-09-23","327,838 npm downloads a week for node-llama-cpp","latest node-llama-cpp@3.22.1"],"log":null},"trust":"unlabeled","health":{"status":"alive","checkedAt":"2026-10-02T20:51:55.833Z","http":200},"description":"[](https://deep-reinforce.com/ornith.html)\n\n# Ornith-1.5-35B-A3B\n\nChirp Chirp! 🐦 We are introducing Ornith-1.5, a major step toward building foundation models through end-to-end self-improvement.\n\nOrnith-1.5 extends Ornith-1.0 (which was developed on top of Qwen3.5 and Gemma4 with additional continued pretraining, mid-training, and post-training) by expanding the self-improvement loop from scaffold and rollout optimization to jointly optimizing task generation, scaffold construction, and solution rollouts. Rather than relying on a fixed set of human-curated tasks and manually designed harnesses, Ornith-1.5 continuously generates new training tasks, discovers effective strategies for solving them, and improves the policy through reinforcement learning. For more details on the task, harness, and rollout reward design, please refer to our [blog](https://ornith.ai/ornith_1_5.html).\n\n## Ornith 1.5 35B-A3B\n\nThis model card documents **Ornith-1.5-35B-A3B**, the mid-size mixture-of-experts member of the Ornith-1.5 family. It activates only ~3B parameters per token, yet significantly outperforms its similar-sized peer Qwen 3.6-35B across all coding and agentic benchmarks, and outperforms dense models such as Gemma 4-31B and Muse Glimmer-30B by wide margins on agentic coding.\n\n### Benchmarks\n\nOrnith-1.5-35B-A3B\nOrnith-1.0-35B-A3B\nQwen3.6-35B-A3B\nGemma-4-31B \nMuse-Glimmer-30B \nQwen3.5-397B\n\nCoding\n\nTerminal-Bench 2.1 (Terminus-2)\n67.8\n64.2\n52.5\n42.1\n51.7\n53.5\n\nTerminal-Bench 2.1 (Claude Code)\n68.5\n62.8\n49.2\n-\n-\n48.6\n\nSWE-bench Verified\n79\n75.6\n73.4\n52\n76\n76.4\n\nSWE-bench Pro\n59.6\n50.4\n49.5\n35.7\n51.2\n51.6\n\nSWE-bench Multilingual\n71.4\n69.3\n67.2\n51.7\n-\n69.3\n\nDeepSWE\n22\n0\n0\n-\n-\n1\n\nFrontier-Bench v0.1\n5.1\n1.4\n1.4\n-\n-\n1.4\n\nNL2Repo\n46.2\n34.6\n29.4\n15.5\n-\n36.8\n\nSWE Atlas - QnA\n39.8\n37.1\n15.5\n-\n-\n20.4\n\nReasoning\n\nHLE (no tools)\n25.6\n20.8\n21.4\n19.5\n22\n28.7\n\nHLE (with tools)\n33.4\n30.1\n28.9\n26.5\n-\n48.3\n\nGPQA Diamond\n89.2\n86…\n\nSource: https://huggingface.co/ornith-ai/Ornith-1.5-35B-A3B-GGUF","install":{"kind":"model","hfId":"ornith-ai/Ornith-1.5-35B-A3B-GGUF","gated":false,"format":"gguf","files":[{"name":"Ornith-1.5-35B-BF16.gguf","size":71066994400,"quant":"BF16","sha256":"044d7f8b55580720dae0d0d8b58f824024901ed5da62f93c011f6bd8dc45a42b"},{"name":"Ornith-1.5-35B-Q4_K_M.gguf","size":21713463040,"quant":"Q4_K_M","sha256":"42739874cc2ccfdb8523b23fbe52e29b2a7555c8176737ca9ca0b5d59859d41f"},{"name":"Ornith-1.5-35B-Q5_K_M.gguf","size":25347532544,"quant":"Q5_K_M","sha256":"91df97de5845100e850b4b5ec5ff35695382020b880fad6f7f51787b3a953bd0"},{"name":"Ornith-1.5-35B-Q6_K.gguf","size":29208731392,"quant":"Q6_K","sha256":"15d4658bbfc9c6034621729c15bbb50662c82b32a7ddd9624a1e545a74bdbb4b"},{"name":"Ornith-1.5-35B-Q8_0.gguf","size":37802149280,"quant":"Q8_0","sha256":"de46c4baf4b4dd85ea438bb0f757f21c38841a353506579979bba114311658c3"},{"name":"mmproj-Ornith-1.5-35B-BF16.gguf","size":902822240,"quant":"BF16","sha256":"1921a36a85aee56cd2abd27f46701802c9d85a33474792e600df6c3b282a135d"}],"totalBytes":186041692896,"suggestedFile":"Ornith-1.5-35B-Q4_K_M.gguf","requirements":{"ramGb":24,"diskBytes":21713463040,"note":"estimate: suggested file size × 1.15 + 0.5 GB; a real measurement comes with lsh models"},"runWith":["llama.cpp","ollama"],"command":"lsh models install hf:ornith-ai/Ornith-1.5-35B-A3B-GGUF"}}