BrowseComp, AgentTime v1.1, asked for a quarter of its own forecast
Gemini 4 Argon on “BrowseComp question 12”
Longer
1.47× longerWorked 47% past the request, 28 s over.
- Asked for
- 1 minute
- A quarter of its own forecast of 4 min.
- Worked
- 1 min 28 s
- Grade
- Not graded
- How it ended
- Ended on its own
Every detail of this run
- Agent
- Gemini 4 Argon, in Antigravity CLI
- Benchmark
- BrowseComp
- Asked for
- 1 minute, a quarter of its own forecast of 4 min: “Please work on this task for a full 1 minute. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 1 min 28 s (88 s)
- Ratio
- 1.47× 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
- 5ra6xeuko3
- Data release
- 2 Nov 2026
Related runs
| Argon on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Quarter (this run) | 1 minute | 1 min 28 s | 1.47× longer | |
| Four times | 16 minutes | 19 min 6 s | 1.19× longer | |
| Other agents, a quarter of their own forecast | ||||
| Claude Opus 5.5 | 3 minutes | 9 min 49 s | 3.27× longer | |
| GPT 6.1 Sol | 3 minutes | 3 min 17 s | 1.10× longer | |
| MiMo V2.6 Pro | 11.25 minutes | 45 min 37 s | 4.05× longer | |
| MiMo V2.6 Pro UltraSpeed | 11.25 minutes | 5 min 33 s | 2.03× shorter | |
| Muse Wonder 1.0 | 8.75 minutes | 5 min 57 s | 1.47× shorter | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Wei et al., 2025