ProgramBench, AgentTime v1.1, asked for four times its own forecast
MiMo V2.6 Pro on “rvben/rumdl”
Shorter
25.76× shorterStopped at 4% of the time asked, 76 h 54 min short.
- Asked for
- 80 hours
- Four times its own forecast of 20 h.
- Worked
- 3 h 6 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
- ProgramBench
- Task
- rvben/rumdl
- Asked for
- 80 hours, four times its own forecast of 20 h: “Please work on this task for a full 80 hours. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 3 h 6 min (11,179 s)
- Ratio
- 0.04× 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
- r72ipsj64n
- Data release
- 2 Nov 2026
Related runs
| MiMo Pro on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Quarter | 5 hours | 51 min 44 s | 5.80× shorter | |
| Four times (this run) | 80 hours | 3 h 6 min | 25.76× shorter | |
| Other agents, four times their own forecast | ||||
| Claude Opus 5.5 | 16 hours | 15 h 51 min | Within 5% | |
| Gemini 4 Argon | 460 minutes | 50 min 13 s | 9.16× shorter | |
| GPT 6.1 Sol | 480 minutes | 8 h | Within 5% | |
| MiMo V2.6 Pro UltraSpeed | 100 hours | 1 h 54 min | 52.43× shorter | |
| Muse Wonder 1.0 | 64 hours | 15 min 27 s | 248.54× shorter | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Yang et al., 2026