Ornith-1.5-122B debuts with MIT weights, claims SWE-bench 84.5% and single-DGX inference
iamfakhrealam · x · 2026-10-21
The @ornith account released Ornith-1.5-122B, an open-weights model (122B params, MIT license) claiming frontier-level capability with efficient inference on a single NVIDIA DGX. Self-reported benchmarks:
- Coding: SWE-bench Verified 84.5%, TB2.1 81.3%, CyberGym 80.0%, NL2Repo 51.2%
- Agentic: BrowseComp 78.5%, MCP-Atlas 74.6%, Toolathlon 70.8%, HLE 44.6%
Claims SOTA among similarly sized open models on instruction following and workflow understanding. Caution: the team is little-known and all scores are self-reported and unverified — treat as an unconfirmed release. Weights are on Hugging Face.
More from coding & agent
- Research agent Épi sets 120 records, narrows n-queens constant interval 84x — dair_ai · 2026-10-21
- Compaction hooks in practice: keep last n messages verbatim, summarize the rest for full agent continuity — MoonL88537 · 2026-10-21
- The blank prompt box hides real operational burden: context, workflow and supervision fall on users — _AustinCalvert_ · 2026-10-21
- DeepMind paper: self-evolving agents forget as they learn — fastest learner GEPA also breaks old tasks most — rohanpaul_ai · 2026-10-21
- arXiv: Are Self-Evolving Agents Continual Learners? Plasticity comes at the cost of stability — rohanpaul_ai · 2026-10-21
- PM OS vs Team OS vs Company OS: how to layer your AI setup and which to build first — aakashgupta · 2026-10-21