{"v":1,"id":"model:hf:ornith-ai/Ornith-1.5-397B-GGUF","slug":"model-ornith-ai-ornith-1-5-397b-gguf","kind":"model","category":"llm","title":"Ornith-1.5-397B-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-1.5-397B-GGUF","url":"https://huggingface.co/ornith-ai/Ornith-1.5-397B-GGUF","rev":"771a73943cafcf88d496c423ea5dd3a1622b1c10","fetchedAt":"2026-10-02T20:52:16.314Z","etag":"W/\"2bfe-Y9CKQCGbnt8eZBiPItb257DvaXc\""},"author":{"name":"ornith-ai","url":"https://huggingface.co/ornith-ai"},"license":{"spdx":"mit","raw":"mit","url":"https://huggingface.co/ornith-ai/Ornith-1.5-397B/blob/main/LICENSE","open":true},"metrics":{"downloads":1260187,"downloadsWeek":327838,"likes":40,"stars":130156,"openIssues":2528,"lastRelease":{"tag":"v0.5.0","at":"2026-09-23T20:50:06Z"},"pushedAt":"2026-10-02T20:17:08Z","takenAt":"2026-10-02T20:52:16.314Z"},"tags":["transformers","gguf","text-generation","endpoints_compatible","imatrix","conversational"],"pipeline":"text-generation","links":{"github":"ggml-org/llama.cpp","npm":"node-llama-cpp"},"updatedAt":"2026-08-24T05:54:49.000Z","collectedAt":"2026-10-02T20:52:16.314Z","review":{"numbers":["1,260,187 downloads on Hugging Face","40 likes","license mit","228 GB for Ornith-1.5-397B-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":"2026-10-02T20:52:16.314Z","http":200},"description":"# Ornith-1.5-397B\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 397B\n\nThis model card documents **Ornith-1.5-397B**, the flagship member of the Ornith-1.5 family — a 397B mixture-of-experts model. It scores 86.1 on Terminal-Bench 2.1 and 56.0 on DeepSWE, performing on par with Claude Opus 4.8 (85.0 and 59.0) while outperforming leading open-source models of similar scale, including GLM-5.2 and DeepSeek-V4-Flash-0731.\n\n### Benchmarks\n\nOrnith-1.5-397B\nDeepSeek-V4-Flash-0731 (284B)\nGLM-5.2 (753B)\nClaude Opus 4.8\nKimi K3 (2.8T)\nOrnith-1.0-397B\n\nCoding\n\nTerminal-Bench 2.1 (Terminus-2)\n86.1\n82.7\n81\n85\n88.3\n77.5\n\nTerminal-Bench 2.1 (Claude Code)\n85.2\n81.8\n82.7\n78.9\n-\n78.2\n\nSWE-bench Verified\n86\n81.6\n83\n85.8\n86.2\n82.4\n\nSWE-bench Pro\n65.1\n64.4\n62.1\n68\n-\n62.2\n\nSWE-bench Multilingual\n79.6\n77.9\n78.4\n75.7\n-\n78.9\n\nDeepSWE\n56\n54.4\n46.2\n59\n67.5\n8\n\nFrontier-Bench v0.1\n13.5\n6.1\n5.1\n21.1\n23\n2.7\n\nNL2Repo\n59.5\n54.2\n48.9\n69.7\n-\n48.2\n\nSWE Atlas - QnA\n55.6\n51.6\n50\n59.7\n59.7\n41.2\n\nReasoning\n\nHLE (no tools)\n44.6\n35\n40.5\n49.8\n43.5\n30.2\n\nHLE (with tools)\n56.1\n50.8\n54.7\n57.9\n56\n47.5\n\nGPQA Diamond\n92.8\n91.4\n91.2\n93.6\n93.5\n88.1\n\nAgentic\n\nMCP-Atlas\n80\n74…\n\nSource: https://huggingface.co/ornith-ai/Ornith-1.5-397B-GGUF","install":{"kind":"model","hfId":"ornith-ai/Ornith-1.5-397B-GGUF","gated":false,"format":"gguf","files":[{"name":"Ornith-1.5-397B-Q4_K_M.gguf","size":244309803808,"quant":"Q4_K_M","sha256":"c7775e6fae1a47619c199c81b865df9014e5d724509a098377ce1c84744b6552"},{"name":"Ornith-1.5-397B-Q5_K_M.gguf","size":286324736800,"quant":"Q5_K_M","sha256":"0154e632f34bd63039df5c15badacb9db1d4cdffa16f8e12aa968eadefbe357e"},{"name":"Ornith-1.5-397B-Q6_K.gguf","size":330965603104,"quant":"Q6_K","sha256":"5ffa1e6b584f3d158023b7f64ad2cb33e6a81cf185c71a3b2d0d7c20dd7372ea"},{"name":"Ornith-1.5-397B-Q8_0.gguf","size":428521390560,"quant":"Q8_0","sha256":"1e033a38f099a5c125e3cf762ef93b6f3db071e18afc55dea6c265c4d0768d8e"},{"name":"mmproj-Ornith-1.5-397B-BF16.gguf","size":921704800,"quant":"BF16","sha256":"9da8c035659d9782b80c2ca6bebdb3befb98a2d7c5a3ac5c001d5a9f02f76fdc"}],"totalBytes":1291043239072,"suggestedFile":"Ornith-1.5-397B-Q4_K_M.gguf","requirements":{"ramGb":263,"diskBytes":244309803808,"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-1.5-397B-GGUF"}}