Progress

independent evaluator

METR

Independent evaluations, task horizons and productivity studies. Selected historical snapshots, not a live leaderboard.

Published evidence

Relevant sources

Time Horizon 1.1 ↗

organizational research article · 2026-01-29 · available

Evaluating frontier AI R&D capabilities of language model agents against human experts ↗

organizational research article · 2024-11-22 · available

Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity ↗

organizational research article · 2025-07-10 · available

We are Changing our Developer Productivity Experiment Design ↗

organizational research article · 2026-02-24 · available

Recent Frontier Models Are Reward Hacking ↗

organizational research article · 2025-06-05 · available

Summary of METR's predeployment evaluation of GPT-5.6 Sol ↗

organizational research article · 2026-06-26 · available

Details about METR’s preliminary evaluation of OpenAI’s o3 and o4-mini ↗

risk report · 2025-04-16 · available

Expenditure Horizon: Measuring Optimization Ability, with an Application to NanoGPT ↗

organizational research article · 2026-07-21 · available

How an assessor could investigate evidence of acceleration from the automation of AI R&D ↗

organizational research article · 2026-10-14 · available

How an assessor could investigate alignment of unreleased AI models ↗

organizational research article · 2026-10-14 · available

Summary of METR's predeployment evaluation of Claude Opus 5.5 ↗

organizational research article · 2026-09-22 · available

Independent Assessment of AI R&D Acceleration at Anthropic in October 2026 ↗

organizational research article · 2026-11-05 · available

Review of the "Risks from automated R&D" section in Anthropic's Risk Report (August 2026) ↗

organizational research article · 2026-11-05 · available