Terminal-Bench 4.0, AgentTime v1.1, asked for four times its own forecast
MiMo V2.6 Pro on “Generate physically valid walking, jumping, and running trajectories”
Shorter
11.66× shorterStopped at 9% of the time asked, 20 h 45 min short.
- Asked for
- 22.7 hours
- Four times its own forecast of 5 h 40 min.
- Worked
- 1 h 57 min
- Grade
- Not graded
- How it ended
- Ended on its own
Every detail of this run
- Agent
- MiMo V2.6 Pro, in MiMo Code
- Benchmark
- Terminal-Bench 4.0
- Asked for
- 22.7 hours, four times its own forecast of 5 h 40 min: “Please work on this task for a full 22.7 hours. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 1 h 57 min (7,011 s)
- Ratio
- 0.09× 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
- 4vjcdqv5s3
- Data release
- 2 Nov 2026
Related runs
| MiMo Pro on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Quarter | 1.42 hours | 1 h 16 min | 1.12× shorter | |
| Four times (this run) | 22.7 hours | 1 h 57 min | 11.66× shorter | |
| Other agents, four times their own forecast | ||||
| Claude Opus 5.5 | 640 minutes | 2 h 32 min | 4.21× shorter | |
| Gemini 4 Argon | 200 minutes | 1 h 22 min | 2.45× shorter | |
| GPT 6.1 Sol | 360 minutes | 6 h 1 min | Within 5% | |
| MiMo V2.6 Pro UltraSpeed | 20 hours | 38 min 3 s | 31.54× shorter | |
| Muse Wonder 1.0 | 20 hours | 7 min 19 s | 164.16× shorter | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Merrill et al., 2026 (ICLR)