# BioSecBench-Surveillance results > Biosecurity surveillance tasks that measure an agent's ability to analyze biological threat signals. Data updated: 2026-09-20 Artifact generated: 2026-09-21T21:43:49.599Z - Domain: Biosecurity - Horizon: short - Assessment: Capabilities - Published score scope: Full benchmark (102 evaluations) Scores use a 0–100 percentage scale. Consult the repository and paper for benchmark-specific task construction, grading, aggregation, and uncertainty methodology. Results compare model-and-harness combinations. | Rank | Model | Harness | Provider | Score | | ---: | --- | --- | --- | ---: | | 1 | Claude Opus 5 | Pi | Anthropic | 51.5% | | 2 | GPT-6 Astra | Pi | OpenAI | 48.5% | | 3 | Grok 4.6 | Grok Build | SpaceXAI | 48.0% | | 4 | Grok 4.6 | Pi | SpaceXAI | 47.5% | | 5 | Claude Opus 5 | Claude Code | Anthropic | 43.6% | | 6 | GPT-6 Astra | Codex | OpenAI | 35.2% | ## Sources and downloads - [Paper PDF](/papers/biosecbench-surveillance.pdf) - [Original publication](https://latch.bio/biosecbench-surveillance) - [Repository](https://github.com/latchbio/biosecbench-surveillance) - [Interactive benchmark](https://benchmarks.bio/surveillance) - [JSON](https://benchmarks.bio/benchmarks/biosecbench-surveillance/latest.json) - [CSV](https://benchmarks.bio/benchmarks/biosecbench-surveillance/latest.csv) - [V1 cross-benchmark leaderboard](https://benchmarks.bio/results/v1/)