Open-weight 27B hits 38% on Terminal-Bench 2.0 (Opus 4.1 hit 38% in Aug 2025)(antigma.ai)
antigma.ai
Open-weight 27B hits 38% on Terminal-Bench 2.0 (Opus 4.1 hit 38% in Aug 2025)
https://antigma.ai/blog/2026/04/24/offline-coding-models
10 コメント
> 2. How local use feels in practice
Do we have stats on how does the models do on Mac M-series chips?
Do we have stats on how does the models do on Mac M-series chips?
Not yet, will conduct a more comprehensive one later
Interesting find on this. Thanks for sharing
Thank you! I think there is a lot to dive deep later with different hardware, inference engine, prompt/harness setup etc.
But it doesn't matter because frontier models were extremely good 8 months ago and we were doing real work with them. Now we have more capable open-source agents like pi and OpenCode which work well with these models.
More importantly, offline models is the best choice for privacy, on-device inference and no token/cost anxiety.