AI research organization Ornith has released Ornith-1.5, a family of open-source large language models. The models come in three sizes: Ornith-1.5-397B, Ornith-1.5-35B-A3B, and Ornith-1.5-9B. The largest model, Ornith-1.5-397B, achieves benchmark scores comparable to Claude Opus 4.8, and a quantized version of the smallest model, Ornith-1.5-9B, is capable of running on smartphones. Ornith-1.5 models were trained using self-improving strategies, incorporating a loop where the AI model itself proposes challenging tasks to enhance its performance autonomously. AI
IMPACT Sets a new benchmark for open-source models, potentially accelerating development and deployment on consumer devices.
RANK_REASON Frontier-lab model release with system card [lever_c_demoted from frontier_release: ic=1 ai=1.0]
Read on Mastodon — mastodon.social →
- Claude Opus 4.8
- DeepSeek-V4-Flash-0731
- Gemma-4-31B
- GLM-5.2
- Muse-Glimmer-30B
- Ornith-1.5
- Ornith-1.5-35B-A3B
- Ornith-1.5-397B
- Ornith-1.5-9B
- Qwen3.5-9B
- Qwen3.6-35B-A3B
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →