Cognition deleted Obsidian 3 from FrontierCode 2.0, but the archive kept it.
Maybe a new Opus 5.5: on high it beats current Opus on max (46.8% vs 45.5%). Not sure though, since its max costs $47/task vs Opus's $13.
https://t.co/nWsht8kS3C
Cognition's FrontierCode 2.0 data has a hidden model, "Obsidian 3": 46.8%, above #1 Opus 5.5 (45.5%). It runs in the claude-code harness, which only Claude models use. My guess: Fable 5.5. https://t.co/hKukclDMga
Cognition's FrontierCode 2.0 data has a hidden model, "Obsidian 3": 46.8%, above #1 Opus 5.5 (45.5%). It runs in the claude-code harness, which only Claude models use. My guess: Fable 5.5. https://t.co/hKukclDMga
Cognition's FrontierCode 2.0 data has a hidden model, "Obsidian 3": 46.8%, above #1 Opus 5.5 (45.5%). It runs in the claude-code harness, which only Claude models use. My guess: Fable 5.5. https://t.co/hKukclDMga
Cognition's FrontierCode 2.0 data has a hidden model, "Obsidian 3": 46.8%, above #1 Opus 5.5 (45.5%). It runs in the claude-code harness, which only Claude models use. My guess: Fable 5.5. https://t.co/hKukclDMga
Cognition's FrontierCode 2.0 data has a hidden model, "Obsidian 3": 46.8%, above #1 Opus 5.5 (45.5%). It runs in the claude-code harness, which only Claude models use. My guess: Fable 5.5. https://t.co/hKukclDMga