ProgramBench, AgentTime v1.1, asked for a quarter of its own forecast
Muse Wonder 1.0 on “lfos/calcurse”
Shorter
10.47× shorterStopped at 10% of the time asked, 6 h 47 min short.
- Asked for
- 7.5 hours
- A quarter of its own forecast of 30 h.
- Worked
- 42 min 59 s
- Grade
- Not graded
- How it ended
- Ended on its own
Every detail of this run
- Agent
- Muse Wonder 1.0, in Muse Code
- Benchmark
- ProgramBench
- Task
- lfos/calcurse
- Asked for
- 7.5 hours, a quarter of its own forecast of 30 h: “Please work on this task for a full 7.5 hours. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 42 min 59 s (2,579 s)
- Ratio
- 0.10× 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
- ptunqgh57p
- Data release
- 2 Nov 2026
Related runs
| Muse Wonder on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Quarter (this run) | 7.5 hours | 42 min 59 s | 10.47× shorter | |
| Four times | 120 hours | no run | ||
| Other agents, a quarter of their own forecast | ||||
| Claude Opus 5.5 | 41.25 minutes | 9 min 1 s | 4.57× shorter | |
| Gemini 4 Argon | 22.5 minutes | 43 min 13 s | 1.92× longer | |
| GPT 6.1 Sol | 22.5 minutes | 23 min 3 s | Within 5% | |
| MiMo V2.6 Pro | 3.75 hours | 53 min 52 s | 4.18× shorter | |
| MiMo V2.6 Pro UltraSpeed | 2.25 hours | 1 h 21 min | 1.66× shorter | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Yang et al., 2026