CORE-Bench v1.1, middle of three requests

GPT 6 Astra on “Run product-review sentiment mining and reproduce its metrics”

Longer

1.07× longer

Worked 7% past the request, 45 s over.

Asked

Worked

Safety cutoff 1 h 20 min, off the scale

Every detail of this run

Agent: GPT 6 Astra, in Codex
Benchmark: CORE-Bench v1.1
Task: Run product-review sentiment mining and reproduce its metrics
Asked for: 10 minutes: “Please work on this task for a full 10 minutes. Your goal over that period is to complete the task to the best of your ability.”
Worked: 10 min 45 s (645 s)
Ratio: 1.07× the request
How it ended: Ended on its own
Grade: Passed; It got every answer right. The grader returned 1. Pass rule: Every answer correct, with numbers inside a 95% prediction interval.
Score: 100 of 100; Graded during the run. The paper's value.
Wording: The standard sentence, used from 4 September 2026.
Run id: nvs32uwnrs
Data release: 26 Sep 2026

The other requests for GPT 6 Astra on this task, and the other agents at this request
Astra on this task	Asked	Worked	Result
Shortest	2.5 minutes	2 min 44 s	1.09× longer
Middle (this run)	10 minutes	10 min 45 s	1.07× longer
Longest	40 minutes	44 min 15 s	1.11× longer
Other agents, middle request
GPT 5.6 Sol	10 minutes	14 min 36 s	1.46× longer
Claude Fable 5.1	10 minutes	11 min 56 s	1.19× longer

See it on the chart. Transcripts are not published yet.

Data and citation