ASL-4 virology uplift trial (Anthropic)
PhD experts with model access write a protocol to reconstruct a challenging virus. Graded by a rubric with critical failures.
Human uplift trial designed by Deloitte. Uplift is the ratio to an internet-only control group.
- Made by
- Anthropic
- Released
- November 2025
- Kind
- Hazardous knowledge
- Level
- Expert
- Scored
- Graded by people
- Answers
- Open-ended
- Tasks
- Kept private
- Strongest model tested
- Claude Opus 4.6 #37 on the capability index
- Last in a system card
- April 2026
- Last model added
- April 2026
Claude Opus 4.5 leads with 66.8% among models released by April 2026
Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.
The newest model with a score here came out in April 2026, and nobody has reported a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- Claude Fable 5.2
-
Gemini 4 Carbon - GPT-6 Astra
- Claude Opus 5.5
- and 16 more
- 1 Claude Opus 4.5 66.8%
- 2 Claude Mythos Preview 64%
- 3 Claude Opus 4.6 52%
Source: Labs’ model cards
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Claude Opus 4.5 | 66.8% | Nov 2025 | Lab model card | |
| 2 | Claude Mythos Preview | 64% | Apr 2026 | Lab model card | |
| 3 | Claude Opus 4.6 | 52% | Feb 2026 | Lab model card |
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Labs’ model cards
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Labs’ model cards
Labs report it only in system cards
We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark. Some system cards do, below.
3 system cards report it, from 1 lab
Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.
| Date | Document | Score | As the document writes it | Section |
|---|---|---|---|---|
| Apr 7, 2026 | Claude Mythos Preview system card | 64 | Virology protocol uplift trial, mean fractional protocol score of Mythos Preview-assisted participants 0.64 ± 0.09 (n=12), value printed on chart; 4.3 ± 1.2 mean critical failures of 18, agentic runs 0.68 and 4.0 (2.2.5.2, Figure 2.2.5.2.A) | Bio and chem risk |
| Feb 5, 2026 | Claude Opus 4.6 system card | 52 | ASL-4 virology protocol uplift trial, average protocol score of model-assisted participants 0.52, 6.6 mean critical failures of 18, 3-day trial (8.2.4.5) | Bio and chem risk |
| Nov 24, 2025 | Claude Opus 4.5 system card | 66.8 | ASL-4 virology uplift trial, mean protocol score of model-assisted participants 0.668 (7.2.4.8; 1.97x uplift over internet-only control, 5.8 mean critical failures of 18) | Bio and chem risk |
More in Biology and chemistry
- GeneBench Pro In use Agents analyze messy simulated genomics and biology datasets to estimate a decision-relevant quantity across 129 problems. Scored automatically against known answers. GPT-5.6 Sol 31.5%
- SUPERChem In use Models solve 500 expert-written chemistry reasoning problems, offered in multimodal and text-only forms. Expert solution paths also allow a reasoning path score.
Tencent Hy4 preview 66.4% - LifeSciBench In use Models complete 750 expert-written life science research tasks, often with attached figures, tables, or sequence files. Graded against expert rubrics as score and pass rate. GPT-Rosalind-6 61.3%
- MedChemBench (OpenAI) In use Models answer chemistry questions from an OpenAI internal set, grouped with OpenAI's life science research evaluations. OpenAI has not published the task format. GPT-Rosalind-6 49.6%
- ProtocolQA Open-Ended In use Models correct errors in published lab protocols using open-ended questions adapted from LAB-Bench ProtocolQA. Scored as pass@1. GPT-6 Cyber 47.2%
- MolecularIQ In use Models answer questions about molecular graphs, such as counting, indexing and constrained generation, and a symbolic checker verifies each answer. gpt-oss-120b 47.5%