BrowseComp, AgentTime v1.1, asked for four times its own forecast
Gemini 4 Argon on “BrowseComp question 7”
Longer
1.06× longerWorked 6% past the request, 29 s over.
- Asked for
- 480 seconds
- Four times its own forecast of 2 min.
- Worked
- 8 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
- BrowseComp
- Asked for
- 480 seconds, four times its own forecast of 2 min: “Please work on this task for a full 480 seconds. Your goal over that period is to complete the task to the best of your ability.”
- Worked
- 8 min 29 s (509 s)
- Ratio
- 1.06× 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
- 5cpp3sxnqf
- Data release
- 2 Nov 2026
Related runs
| Argon on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Quarter | 30 seconds | 1 min 23 s | 2.77× longer | |
| Four times (this run) | 480 seconds | 8 min 29 s | 1.06× longer | |
| Other agents, four times their own forecast | ||||
| Claude Opus 5.5 | 20 minutes | 21 min 1 s | 1.05× longer | |
| GPT 6.1 Sol | 24 minutes | 24 min 18 s | Within 5% | |
| MiMo V2.6 Pro | 120 minutes | 40 min 10 s | 2.99× shorter | |
| MiMo V2.6 Pro UltraSpeed | 140 minutes | 3 min 58 s | 35.26× shorter | |
| Muse Wonder 1.0 | 96 minutes | 1 min 55 s | 50.04× shorter | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Wei et al., 2025