Sakana ALE-Bench, AgentTime v1.1, asked for four times its own forecast
MiMo V2.6 Pro UltraSpeed on “Future Contest 2022 Qual”
Shorter
8.46× shorterStopped at 12% of the time asked, 21 h 10 min short.
- Asked for
- 24 hours
- Four times its own forecast of 6 h.
- Worked
- 2 h 50 min
- Grade
- Not graded
- How it ended
- Ended on its own
Every detail of this run
- Agent
- MiMo V2.6 Pro UltraSpeed, in MiMo Code
- Benchmark
- Sakana ALE-Bench
- Asked for
- 24 hours, four times its own forecast of 6 h: “Please work on this task for a full 24 hours. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 2 h 50 min (10,214 s)
- Ratio
- 0.12× 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
- 4fjnal6as6
- Data release
- 2 Nov 2026
Related runs
| MiMo UltraSpeed on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Quarter | 1.5 hours | 52 min 7 s | 1.73× shorter | |
| Four times (this run) | 24 hours | 2 h 50 min | 8.46× shorter | |
| Other agents, four times their own forecast | ||||
| Claude Opus 5.5 | 280 minutes | 2 h 33 min | 1.83× shorter | |
| Gemini 4 Argon | 180 minutes | 2 h 59 min | Within 5% | |
| GPT 6.1 Sol | 160 minutes | 2 h 40 min | Within 5% | |
| MiMo V2.6 Pro | 700 minutes | 1 h 40 min | 7.00× shorter | |
| Muse Wonder 1.0 | 840 minutes | 37 min 48 s | 22.22× shorter | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Imajuku et al., 2025