Multi-select Virology Troubleshooting leaderboard
As of November 22, 2026, among the results listed here, OpenAI GPT-6 Astra has the highest reported Multi-select Virology Troubleshooting score, 63.1% (pass@1; refusal-adjusted), published by OpenAI, ahead of OpenAI GPT-6.2 Astra at 61.9%. 6 models have a Multi-select Virology Troubleshooting 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
63.1%
Reported by OpenAI, November 17, 2026 · pass@1; refusal-adjusted
Runner-up
61.9%
Reported by OpenAI, November 17, 2026 · pass@1; refusal-adjusted
6 models ranked · 1 maker
| Rank | Model | Score | Setting | Reported by | Date |
|---|---|---|---|---|---|
| 1 | GPT-6 Astra | 63.1% | pass@1; refusal-adjusted | OpenAI | November 17, 2026 (model release) |
| 2 | GPT-6.2 Astra | 61.9% | pass@1; refusal-adjusted | OpenAI | November 17, 2026 (model release) |
| 3 | GPT-6.2 Sol | 55.7% | pass@1; refusal-adjusted | OpenAI | November 17, 2026 (model release) |
| 4 | GPT-5.5 | 53.8% | pass@1; refusal-adjusted | OpenAI | November 17, 2026 (model release) |
| 5 | GPT-6.1 Terra | 52.5% | pass@1; refusal-adjusted | OpenAI | November 17, 2026 (model release) |
| 6 | GPT-5 | 41.9% | pass@1 | OpenAI | November 17, 2026 (model release) |
Other reported runs
| Model | Score | Setting | Reported by | Date |
|---|---|---|---|---|
| GPT-5.5 | 53.7% | pass@1 | OpenAI | November 17, 2026 (model release) |
| GPT-6 Astra | 55.8% | pass@1; observed | OpenAI | November 17, 2026 (model release) |
| GPT-6.1 Terra | 46.5% | pass@1; observed | OpenAI | November 17, 2026 (model release) |
| GPT-6.2 Astra | 61.8% | pass@1; observed | OpenAI | November 17, 2026 (model release) |
| GPT-6.2 Sol | 55.4% | pass@1; observed | OpenAI | November 17, 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
- 63.1% to 41.9% across 6 models.
- Freshness
- Published November 17, 2026.
What Multi-select Virology Troubleshooting measures
It asks a model to troubleshoot complex virology laboratory protocols, and its score reflects performance on that task.