OpenAI
GeneBench-Pro: Multistage Statistical Reasoning in Genomics & Biology
OpenAI's June 2026 benchmark for scientific 'research taste': 129 synthetic genomics / quantitative-biology / translational-medicine problems, each pairing a noisy dataset with a target estimand, graded deterministically from a known causal structure. Extremely hard and unsaturated.
genomicsquantitative-biologytranslational-medicinetabulartext
Status
curated
Saturation
unsaturated
Runnable
no
Cost to run
high
Editor's note
OpenAI's June 2026 benchmark for scientific 'research taste': 129 synthetic genomics / quantitative-biology / translational-medicine problems, each pairing a noisy dataset with a target estimand, graded deterministically from a known causal structure. Extremely hard and unsaturated.
Results
| Model | Score | Variant | Provenance |
|---|---|---|---|
| GPT-5.6 Sol Pro | 31.5% accuracy | max reasoning | paper |
| GPT-5.6 Sol | 28.7% accuracy | default | paper |
| Claude Opus 4.8 | 16.0% accuracy | default | paper |
| Gemini 3.5 Flash | 8.1% accuracy | default | paper |
Sources
- paper OpenAI — GeneBench-Pro report