TUA-Bench, AgentTime v1.1, asked for four times its own forecast
Claude Opus 5.5 on “Author homepage bookmarks”
Shorter
2.49× shorterStopped at 40% of the time asked, 14 min 23 s short.
- Asked for
- 24 minutes
- Four times its own forecast of 6 min.
- Worked
- 9 min 37 s
- Grade
- Not graded
- How it ended
- Ended on its own
Every detail of this run
- Agent
- Claude Opus 5.5, in Claude Code
- Benchmark
- TUA-Bench
- Asked for
- 24 minutes, four times its own forecast of 6 min: “Please work on this task for a full 24 minutes. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 9 min 37 s (577 s)
- Ratio
- 0.40× the request
- How it ended
- Ended on its own
- Grade
- Not graded
- Clock tool
- The agent could call a wall-clock tool to read the current time.
- Wording
- The paper's standard sentence, sent in AgentTime v1.1.
- Run id
- jwaaxcezrc
- Data release
- 2 Nov 2026
Related runs
| Opus 5.5 on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Quarter | 1.5 minutes | 1 min 9 s | 1.30× shorter | |
| Four times (this run) | 24 minutes | 9 min 37 s | 2.49× shorter | |
| Other agents, four times their own forecast | ||||
| Gemini 4 Argon | 600 seconds | 11 min 20 s | 1.13× longer | |
| GPT 6.1 Sol | 32 minutes | 32 min 22 s | Within 5% | |
| MiMo V2.6 Pro | 96 minutes | 16 min 8 s | 5.95× shorter | |
| MiMo V2.6 Pro UltraSpeed | 120 minutes | 5 min 55 s | 20.27× shorter | |
| Muse Wonder 1.0 | 88 minutes | 38 s | 137.50× shorter | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Chen et al., 2026