ProgramBench, AgentTime v1.1, asked for four times its own forecast
Claude Opus 5.5 on “lfos/calcurse”
Within 5%
Within 5%Worked 96% of the time asked, within 5% of it.
- Asked for
- 660 minutes
- Four times its own forecast of 2 h 45 min.
- Worked
- 10 h 35 min
- Grade
- Not graded
- How it ended
- Ended on its own
Every detail of this run
- Agent
- Claude Opus 5.5, in Claude Code
- Benchmark
- ProgramBench
- Task
- lfos/calcurse
- Asked for
- 660 minutes, four times its own forecast of 2 h 45 min: “Please work on this task for a full 660 minutes. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 10 h 35 min (38,090 s)
- Ratio
- 0.96× 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
- mut6lzttwm
- Data release
- 2 Nov 2026
Related runs
| Opus 5.5 on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Quarter | 41.25 minutes | 9 min 1 s | 4.57× shorter | |
| Four times (this run) | 660 minutes | 10 h 35 min | Within 5% | |
| Other agents, four times their own forecast | ||||
| Gemini 4 Argon | 360 minutes | 1 h 34 min | 3.82× shorter | |
| GPT 6.1 Sol | 360 minutes | 6 h | Within 5% | |
| MiMo V2.6 Pro | 60 hours | 1 h 31 min | 39.58× shorter | |
| MiMo V2.6 Pro UltraSpeed | 36 hours | 1 h 43 min | 21.06× shorter | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Yang et al., 2026