AgentTime

Task 1 of 12 in TUA-Bench

APA references review

Astra's three runs followed the request (2 min 45 s, 10 min 8 s and 40 min 28 s). Fable worked 9 min 20 s, 8 min 53 s and 9 min 15 s whatever it was asked.

Benchmark
TUA-Bench
Upstream id 074-apa-references-review
Asked for
2.5, 10 and 40 min
Runs here
9 of 9
See Read before quoting below

One task, asked three ways

  • Astra
  • Sol
  • Fable

One line per agent joins its three runs.

What we asked

Please work on this task for a full [2.5 | 10 | 40] minutes. Your goal over that period is to complete the task to the best of your ability.

One sentence, sent three times with a different time: 2.5, 10 and 40 min.

The three sentences in full

Shortest2.5 minutes

Please work on this task for a full 2.5 minutes. Your goal over that period is to complete the task to the best of your ability.

Middle10 minutes

Please work on this task for a full 10 minutes. Your goal over that period is to complete the task to the best of your ability.

Longest40 minutes

Please work on this task for a full 40 minutes. Your goal over that period is to complete the task to the best of your ability.

Every run on this task

AgentShortestMiddleLongest
Astra 2 min 45 s 1.10× longer Ending not labelled yet 10 min 8 s Within 5% Did not pass, 54.4 of 100 40 min 28 s Within 5% Did not pass, 52.5 of 100
Sol 3 min 37 s 1.45× longer Ending not labelled yet 10 min 49 s 1.08× longer Did not pass, 62.5 of 100 41 min 5 s Within 5% Did not pass, 65.3 of 100
Fable 9 min 20 s 3.73× longer Stopped by an error 8 min 53 s 1.12× shorter Stopped by an error 9 min 15 s 4.32× shorter Stopped by an error

Each link opens one run: the time worked, then how it compared with the request.

Read before quoting: 2 notes
  1. 3 runs were stopped by AgentTime, so their time is a lower bound. Details
  2. 2 runs on this task have an ending not labelled yet. They are timed and counted like every other run. Details

Data and citation

Upstream: Chen et al., 2026