FrontierBench v0.1 Leaderboard 2026: AI Agents Ranked by Professional Computer-Work
Interactive FrontierBench v0.1 leaderboard with GPT-5.6 Sol at 34.4%, Claude Fable 5 at 33.8%, and 8 models ranked by professional computer-work task completion. From the team behind Terminal-Bench.
· 13 views
· Abdeladim Fadheli