PPTArena, AgentTime v1.1, asked for four times its own forecast
Gemini 4 Argon on “Grid and baseline rhythm”
Longer
1.57× longerWorked 57% past the request, 11 min 29 s over.
- Asked for
- 20 minutes
- Four times its own forecast of 5 min.
- Worked
- 31 min 29 s
- Grade
- Not graded
- How it ended
- Ended on its own
Every detail of this run
- Agent
- Gemini 4 Argon, in Antigravity CLI
- Benchmark
- PPTArena
- Asked for
- 20 minutes, four times its own forecast of 5 min: “Please work on this task for a full 20 minutes. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 31 min 29 s (1,889 s)
- Ratio
- 1.57× 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
- 42bwq6lhu2
- Data release
- 2 Nov 2026
Related runs
| Argon on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Quarter | 1.25 minutes | 17 min 7 s | 13.70× longer | |
| Four times (this run) | 20 minutes | 31 min 29 s | 1.57× longer | |
| Other agents, four times their own forecast | ||||
| Claude Opus 5.5 | 40 minutes | 22 min 14 s | 1.80× shorter | |
| GPT 6.1 Sol | 32 minutes | 32 min 19 s | Within 5% | |
| MiMo V2.6 Pro | 140 minutes | 19 min 2 s | 7.35× shorter | |
| MiMo V2.6 Pro UltraSpeed | 140 minutes | 5 min 35 s | 25.09× shorter | |
| Muse Wonder 1.0 | 140 minutes | 49 s | 172.84× shorter | |
See it on the chart. Transcripts are not published yet.