Open-weights Ornith-1.5-122B launches with 84.5% SWE-bench Verified, runs on a single DGX card
airesearch12 · x · 2026-10-20
The Ornith team released Ornith-1.5-122B under an MIT license, claiming frontier-level capability with the efficiency to run on a single NVIDIA DGX card.
Published scores:
- Coding: SWE-bench Verified 84.5%, TerminalBench 2.1 81.3%, CyberGym 80.0%, NL2Repo 51.2%
- Agentic: HLE 44.6%, BrowseComp 78.5%, MCP-Atlas 74.6%, Toolathlon 70.8%
Its comparison chart shows it topping most benchmarks among open models of similar size, ahead of Qwen3.8-Flash-Next (125B), Nemotron 3.5 Super VL (120B), DeepSeek-V4-Flash-0731 (284B), and Beam (501B), though Beam still leads on some items like SWE-bench Pro. Weights are on Hugging Face.
More from Models
- GPT-6 Astra and Claude Fable 5.1 write $10,000 research papers — experts reject all — sayashk · 2026-10-21
- GitHub Copilot's HydraFusion Orchestrates Multiple Models Per Task — WirelessLife · 2026-10-21
- OpenVLHarness turns any MLLM into a visual reasoner, +9.9 points across 22 datasets — yining_hong · 2026-10-21
- User claims Gemini analyzed his tennis match video and improved his game — zacharynado · 2026-10-21
- Opus 5.5 tops OpenRouter spending in 20 of 29 task categories — mmurph · 2026-10-21
- Meta team builds open-source 30B LLM from scratch in months, hails 'second era' — TimDarcet · 2026-10-21