Published full-model reference, then measured local recipes.
| Recipe | GPQA | Terminal | Verified |
|---|---|---|---|
| OpenAI comparison table / publishedPublished reference | 94.3 | 70.7 | — |
The published full-model reference is the top row — or the dashed gold tick on Graphic — not rank #1. A blank cell, missing tick, or missing bead has no recorded cite. Scores are never invented.
ON THIS SETUP · SOURCE-REPORTED
Cloud / API · Reported evaluation
Multi-step tasks in a terminal environment.
Version 2.1 · sample size not reported
This is the source’s reported aggregate. Its metric, agent configuration, and evaluation protocol determine how it should be interpreted. Read the original grading definition before comparing completion rates across setups.
Graduate-level science questions.
Diamond · version 1.0 · sample size not reported
This is the source’s reported aggregate. Its metric, agent configuration, and evaluation protocol determine how it should be interpreted. Read the original grading definition before comparing completion rates across setups.
Task traces are not available for this setup.
Context, sampling, harness settings and dataset versions can vary by benchmark. Each evidence link above contains its run-specific configuration.