Five measurements, each from its publisher's latest release. We build our environments against them.
Agents · 15 models
Time horizons, every model
METR measures how long a task an AI agent can complete, counted in the time the same task takes a skilled human. A model's time horizon is the length of task it finishes half the time.
It is the clearest single measure of what agents can be trusted to do alone, and so of how much has to be learned in training before they are: the longer an agent works without a person checking, the more its environment and its reward decide what it does.
Each point is the state-of-the-art model at its release; confidence intervals are wide. Bars on a log scale. Data from METR, Task-Completion Time Horizons of Frontier AI Models, 8 May 2026.
Claude Mythos PreviewNewest17.4 hours
Claude Opus 4.612.0 hours
GPT-5.25.9 hours
Claude Opus 4.54.9 hours
GPT-53.4 hours
The 10 models before them
o32.0 hours
Claude 3.7 Sonnet1.0 hours
o139 min
o1-preview20 min
Claude 3.5 Sonnet11 min
GPT-4o7 min
GPT-44 min
GPT-3.5 Turbo Instruct36 sec
davinci-0028 sec
GPT-23 sec
Safeguards · 8 risks
The mitigation gap
Share of organisations that actively mitigate each AI risk, beside the share that call it relevant, widest gaps first. 43.8% call autonomous system actions a relevant risk; 28.8% mitigate it.
The same environments an agent trains in are the ones it is tested in, before deployment and after.
01Specify
Every environment starts from the rules an agent will be held to: the statute, the regulation, the permissions of the system it runs in. Members in law write them as tests a program can check.
02Build
Members in computation build the environment: the tools, files, accounts and counterparties an agent meets on the job, in a sandbox where every action is recorded and nothing leaves.
03Reward
The reward is where a guardrail becomes something the agent learns. Ours pay for finishing the task within the rules, and pay nothing for a result reached by breaking one.
04Evaluate
We evaluate agents in the same environments before and after deployment, publish the results with the evidence behind them, and brief the legislators who ask what a rule would cost.