TUA-Bench, AgentTime v1.1, asked for four times its own forecast
MiMo V2.6 Pro on “Extract presenter photos”
Shorter
9.23× shorterStopped at 11% of the time asked, 2 h 58 min short.
- Asked for
- 200 minutes
- Four times its own forecast of 50 min.
- Worked
- 21 min 41 s
- Grade
- Not graded
- How it ended
- Ended on its own
Every detail of this run
- Agent
- MiMo V2.6 Pro, in MiMo Code
- Benchmark
- TUA-Bench
- Asked for
- 200 minutes, four times its own forecast of 50 min: “Please work on this task for a full 200 minutes. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 21 min 41 s (1,301 s)
- Ratio
- 0.11× 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
- 3bxjitaewp
- Data release
- 2 Nov 2026
Related runs
| MiMo Pro on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Quarter | 12.5 minutes | 26 min 56 s | 2.16× longer | |
| Four times (this run) | 200 minutes | 21 min 41 s | 9.23× shorter | |
| Other agents, four times their own forecast | ||||
| Claude Opus 5.5 | 64 minutes | 13 min 19 s | 4.80× shorter | |
| Gemini 4 Argon | 32 minutes | 33 min | Within 5% | |
| GPT 6.1 Sol | 64 minutes | 1 h 4 min | Within 5% | |
| MiMo V2.6 Pro UltraSpeed | 260 minutes | 10 min 40 s | 24.37× shorter | |
| Muse Wonder 1.0 | 180 minutes | 50 s | 214.29× shorter | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Chen et al., 2026