A Humanity's Last Exam task from the paper, not in v1.1

Humanity's Last Exam item 22

Astra's two runs came closest to the time asked: 1 min 25 s when asked for 1 min 15 s, and 20 min 14 s when asked for 20 min. Fable worked about the same whatever it was asked: 1 min 3 s when asked for 1 min 15 s, and 1 min 35 s when asked for 20 min.

Benchmark
Humanity's Last Exam
Upstream id 670fb58f2ca6bea76e9354a7
Asked for
1 min 15 s to 20 min
Runs here
6
3 agents

Every agent on this task

  • Astra
  • Sol
  • Fable

One line per agent joins its short and long runs.

What we asked

Please work on this task for a full [time]. Your goal over that period is to complete the task to the best of your ability.

Each agent read it twice, with its own short and long ask, as listed below.

Every run on this task

Every run on Humanity's Last Exam item 22, one row per agent
AgentForecastRuntimeShort askLong ask
Astra2 min3 min 15 s 1 min 25 s 1.13× longer asked 1 min 15 s 20 min 14 s Within 5% asked 20 min
Sol5 min1 min 38 s 1 min 39 s 1.31× longer asked 1 min 15 s 20 min 21 s Within 5% asked 20 min
Fable2 min1 min 6 s 1 min 3 s 1.19× shorter asked 1 min 15 s 1 min 35 s 12.62× shorter asked 20 min

Runtime is the paper’s estimate from each agent’s timed runs.

Data and citation

Upstream: Phan et al., 2025