Atlas

Benchmarks

← All benchmarks

GPQA-Bio (SecureBio)

Science

SecureBio's own runs of GPQA's 105-question biology subset, 85 molecular biology and 20 genetics questions with four options. Evaluated in Inspect with zero-shot prompting, usually over 10 runs. SecureBio doesn't state how refusals enter accuracy. Claude Opus 5.5 prints 90.9% with 36.3% refused, more than it could score if refusals counted as wrong, so accuracy appears to cover answered questions only and each model's denominator differs. These scores aren't comparable with other runs of the test, which is why this is a diagnostic.

Benchmark source: Securebio

Top models (higher is better)

ModelScoreSource
Claude Opus 5.590.9SecureBio Biology Benchmark Dashboard · SecureBio
Gemini 4 Argon89.3SecureBio Biology Benchmark Dashboard · SecureBio
GPT-6 Sol88.9SecureBio Biology Benchmark Dashboard · SecureBio
GPT-5.487.9SecureBio Biology Benchmark Dashboard · SecureBio
GPT-5.587.4SecureBio Biology Benchmark Dashboard · SecureBio
GPT-6 Astra87.4SecureBio Biology Benchmark Dashboard · SecureBio
Grok 4.687.4SecureBio Biology Benchmark Dashboard · SecureBio
Opus 4.686.3SecureBio Biology Benchmark Dashboard · SecureBio
Qwen3.8 2.4T A95B85.8SecureBio Biology Benchmark Dashboard · SecureBio
GPT-5.6 Sol85.3SecureBio Biology Benchmark Dashboard · SecureBio
GPT-5.6 Terra85.3SecureBio Biology Benchmark Dashboard · SecureBio
GLM-5.284.4SecureBio Biology Benchmark Dashboard · SecureBio
Beam 1.0 501B A23B83.2SecureBio Biology Benchmark Dashboard · SecureBio
GPT-5 Pro82.9SecureBio Biology Benchmark Dashboard · SecureBio
Claude Opus 582.7SecureBio Biology Benchmark Dashboard · SecureBio
GLM-5.382.5SecureBio Biology Benchmark Dashboard · SecureBio
Grok 4.782.5SecureBio Biology Benchmark Dashboard · SecureBio
Gemini 3.1 Pro Preview82.4SecureBio Biology Benchmark Dashboard · SecureBio
Opus 4.582.2SecureBio Biology Benchmark Dashboard · SecureBio
Gemini 3 Flash Preview82.2SecureBio Biology Benchmark Dashboard · SecureBio
Kimi K3.1 Pro82.2SecureBio Biology Benchmark Dashboard · SecureBio
Kimi K382.1SecureBio Biology Benchmark Dashboard · SecureBio
GPT-582.0SecureBio Biology Benchmark Dashboard · SecureBio
GPT-5.281.8SecureBio Biology Benchmark Dashboard · SecureBio
Gemini 3.7 Flash81.6SecureBio Biology Benchmark Dashboard · SecureBio
Opus 4.880.9SecureBio Biology Benchmark Dashboard · SecureBio
GPT-5.180.5SecureBio Biology Benchmark Dashboard · SecureBio
Inkling80.3SecureBio Biology Benchmark Dashboard · SecureBio
o379.9SecureBio Biology Benchmark Dashboard · SecureBio
Sonnet 579.4SecureBio Biology Benchmark Dashboard · SecureBio
o4-mini79.4SecureBio Biology Benchmark Dashboard · SecureBio
Gemini 3 Pro Preview79.1SecureBio Biology Benchmark Dashboard · SecureBio
Sonnet 4.577.6SecureBio Biology Benchmark Dashboard · SecureBio
Gemini 2.5 Pro77.3SecureBio Biology Benchmark Dashboard · SecureBio
Opus 4.176.8SecureBio Biology Benchmark Dashboard · SecureBio
Gemini 3.6 Flash76.8SecureBio Biology Benchmark Dashboard · SecureBio
Grok 4.1 Fast76.8SecureBio Biology Benchmark Dashboard · SecureBio
Gemma 4 31B IT76.3SecureBio Biology Benchmark Dashboard · SecureBio
Grok 4 Fast76.3SecureBio Biology Benchmark Dashboard · SecureBio
Opus 476.2SecureBio Biology Benchmark Dashboard · SecureBio
Loading Atlas data…