AAV leaderboard
As of November 22, 2026, among the results listed here, OpenAI GPT-6.2 Astra has the highest reported AAV score, 0.6 (Mean Spearman rank correlation (higher is better)), published by OpenAI, ahead of OpenAI GPT-6 Astra at 0.6. 9 models have a AAV result in our records.
Scores as the model makers published them, not re-run by us; each row links to its announcement.
Best reported score
0.6
Reported by OpenAI, November 17, 2026 · Mean Spearman rank correlation (higher is better)
Runner-up
0.6
Reported by OpenAI, November 17, 2026 · Mean Spearman rank correlation (higher is better)
9 models ranked · 1 maker
| Rank | Model | Score | Setting | Reported by | Date |
|---|---|---|---|---|---|
| 1 | GPT-6.2 Astra | 0.6 | Mean Spearman rank correlation (higher is better) | OpenAI | November 17, 2026 (model release) |
| 2 | GPT-6 Astra | 0.6 | Mean Spearman rank correlation (higher is better) | OpenAI | November 17, 2026 (model release) |
| 3 | GPT-6.2 Sol | 0.6 | Mean Spearman rank correlation (higher is better) | OpenAI | November 17, 2026 (model release) |
| 4 | GPT-5.5 | 0.6 | Mean Spearman rank correlation (higher is better) | OpenAI | November 17, 2026 (model release) |
| 5 | GPT-6 Sol | 0.5 | Spearman rank correlation (higher is better) | OpenAI | September 4, 2026 (model release) |
| 6 | GPT-5.6 Sol | 0.5 | Spearman rank correlation (higher is better) | OpenAI | September 4, 2026 (model release) |
| 7 | GPT-6.1 Terra | 0.5 | Spearman rank correlation (higher is better) | OpenAI | October 26, 2026 (model release) |
| 8 | GPT-6.1 Sol | 0.5 | Spearman rank correlation (higher is better) | OpenAI | September 4, 2026 (model release) |
| 9 | GPT-5 | 0.5 | Mean Spearman rank correlation (higher is better) | OpenAI | November 17, 2026 (model release) |
Other reported runs
| Model | Score | Setting | Reported by | Date |
|---|---|---|---|---|
| GPT-6 Astra | 0.5 | Spearman rank correlation (higher is better) | OpenAI | September 4, 2026 (model release) |
| GPT-6 Astra | 0.5 | Spearman rank correlation (higher is better) | OpenAI | October 26, 2026 (model release) |
How much to trust these numbers
- Source
- Reported by the model makers themselves. Not re-run by LLMBenchmark or anyone independent.
- Settings
- Settings differ between makers. Check the Setting column before comparing two scores.
- Spread
- 0.6 to 0.5 across 9 models.
- Freshness
- Newest result published November 17, 2026; oldest September 4, 2026.
What AAV measures
AAV tests designing a virus capsid that reaches a target organ and can be manufactured; the results do not explain what its score means.