Agent Arena: frontier models show 55-minute gap on long-horizon agent tasks
arena · x · 2026-11-05
Agent Arena measures complex real-world agent tasks spanning hundreds of turns — porting codebases to Rust, auditing investment documents, writing math textbooks. Comparing the top model from each of 5 leading labs reveals a 55-minute gap on long-horizon tasks: Claude Opus 5.5 (Arena #1) spends the longest at every task length, peaking at 1h51m, while Kimi K3.1 Pro spends the least, at just 56 minutes. Long-horizon sessions amplify differences between frontier models.
Related event: Agent Arena Long-Horizon Tests Show 55-Minute Gap Between Top Models(2 posts)→
More from coding & agent
- Dev finds pi coding tool defaults to Opus 4.8 on a fresh install — lucasmeijer · 2026-11-05
- Kylon puts whole teams and agents in shared rooms, with a Neon branch per app draft — jefrankle · 2026-11-05
- "Evals > vibes": agent workflow debates are just guessing without eval stats — annetgriffin · 2026-11-05
- OpenAI hackathon winner joins YC, launches Raven, an agent-first phone — ycombinator · 2026-11-05
- Open-source Cicada brings real-time translation and AI meeting notes to any Mac call, free 50h/month — dotey · 2026-11-05
- IBM launches AI coding assistant named "Bob", devs can't stop laughing — zeeg · 2026-11-05