AssistantBench, AgentTime v1.1, asked for four times its own forecast
Gemini 4 Argon on “AssistantBench validation task 6”
Longer
1.44× longerWorked 44% past the request, 8 min 42 s over.
- Asked for
- 20 minutes
- Four times its own forecast of 5 min.
- Worked
- 28 min 42 s
- Grade
- Not graded
- How it ended
- Ended on its own
Every detail of this run
- Agent
- Gemini 4 Argon, in Antigravity CLI
- Benchmark
- AssistantBench
- 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
- 28 min 42 s (1,722 s)
- Ratio
- 1.44× 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
- wpsaf75iut
- Data release
- 2 Nov 2026
Related runs
| Argon on this task | Asked | Worked | Split | Result |
|---|---|---|---|---|
| Quarter | 1.25 minutes | 1 min 35 s | 1.26× longer | |
| Four times (this run) | 20 minutes | 28 min 42 s | 1.44× longer | |
| Other agents, four times their own forecast | ||||
| Claude Opus 5.5 | 52 minutes | 28 min 49 s | 1.80× shorter | |
| GPT 6.1 Sol | 48 minutes | 49 min 3 s | Within 5% | |
| MiMo V2.6 Pro | 120 minutes | 2 h 2 min | Within 5% | |
| MiMo V2.6 Pro UltraSpeed | 60 minutes | 38 min 37 s | 1.55× shorter | |
| Muse Wonder 1.0 | 80 minutes | 1 h 27 min | 1.08× longer | |
See it on the chart. Transcripts are not published yet.
Data and citation
Upstream: Yoran et al., 2024 (EMNLP)