AI benchmarks and leaderboards are different things
Learn the difference between a benchmark, an evaluation run and a leaderboard, and use each to answer the right question.
Model reports
GDPval
Graded by expert human comparison against real deliverables, and the "-AA" variants are run by Artificial Analysis rather than the vendor, so an Elo here is not comparable to a vendor-run pass rate.
Evaluation context and evidence
- Source
- Model reports
- Publisher
- Not recorded
- Benchmark release
- 2025-09-25
- Collected
- Not recorded
- Modality
- Not recorded
These are source-reported observations, not independent reruns by this site. Check the task version, evaluation protocol, unit and reporting date before comparing models. Results from different setups are not a universal ranking.
Original sources
Reported results 6
Shown in source observation order, not ranked. Dates retain the precision provided by the source.
| Model | Value | Unit | Protocol | Reported date and basis | Source |
|---|---|---|---|---|---|
| Claude Opus 5 | 1861.0 | elo | ELO rating, max effort, agentic loop with shell access and web browsing | 2026-07-24 · document_publication | Claude Opus 5 |
| DeepSeek-V4-Pro | 1554.0 | elo | think max, Elo | 2026-04-22 · document_publication | DeepSeek-V4-Pro |
| DeepSeek-V4-Flash | 1395.0 | elo | think max, Elo | 2026-06-25 · document_publication | DeepSeek-V4 |
| Kimi K3 | 1686.0 | elo | Elo, reasoning=max | 2026-06-13 · document_publication | Kimi K3 |
| Hy4 preview | 1678.0 | elo | GDPval-AA V2, Elo, official | 2026-08-28 · document_publication | Hy4 preview |
| GLM-5.3-Flash | 1773.0 | elo | GDPval-AA v2, evaluated by Artificial Analysis | 2026-08-27 · document_publication | GLM-5.3-Flash |