MedChemBench (OpenAI)
Models answer chemistry questions from an OpenAI internal set, grouped with OpenAI's life science research evaluations. OpenAI has not published the task format.
Reported in the GPT-5.6 post (Sol 48.3%) and the GPT-6 Astra post (Astra 49.3%). OpenAI omits Claude Fable 5 and 5.1 because they refuse most questions. Format and scoring are guesses.
- Made by
- OpenAI
- Released
- July 2026
- Kind
- Genomics and life sciences
- Level
- Expert
- Scored
- Checked automatically
- Answers
- Short answer
- Tasks
- Kept private
- Strongest model tested
- GPT-6 Astra #3 on the capability index
- Last in a launch post
- September 2026
- Last model added
- October 2026
GPT-Rosalind-6 leads with 49.6%
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.
- 1 GPT-Rosalind-6 49.6%
- 2 GPT-6 Astra 49.3%
- 3 GPT-5.6 Sol 48.3%
- 4 GPT-5.6 Terra 35%
- 5 GPT-5.6 Luna 30.4%
Source: Labs’ model cards
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | GPT-Rosalind-6 | 49.6% | Oct 2026 | Lab model card | |
| 2 | GPT-6 Astra | 49.3% | Sep 2026 | Lab model card | |
| 3 | GPT-5.6 Sol | 48.3% | Jul 2026 | Lab model card | |
| 4 | GPT-5.6 Terra | 35% | Jul 2026 | Lab model card | |
| 5 | GPT-5.6 Luna | 30.4% | Jul 2026 | Lab model card |
3 models have held number one in labs’ own numbers since its release in July 2026
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
One lab put it in a launch post in the last six months
Launch posts from every lab that has listed it, one tick per post. Tall colored ticks include it, and short gray ones leave it out.
| Date | Launch post | Score | As the post writes it |
|---|---|---|---|
| Sep 3, 2026 | GPT-6 Astra | 49.3 | MedChemBench (Internal) |
| Jul 9, 2026 | GPT-5.6 Luna | 30.4 | MedChemBench (Internal) |
| Jul 9, 2026 | GPT-5.6 Terra | 35 | MedChemBench (Internal) |
| Jul 9, 2026 | GPT-5.6 Sol | 48.3 | MedChemBench (Internal) |
One system card reports it
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 |
|---|---|---|---|---|
| Oct 30, 2026 | GPT-Rosalind-6 system card | 49.6 | MedChemBench v7, highest reasoning budget, mean rubric score | Capabilities |
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%
- 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%
- TOMG-Bench In use Models generate molecules from natural language requests in three tasks: molecule editing, molecule optimization and customized generation. Each output is checked for validity and for meeting the requested properties. GPT-3.5 Turbo 18.6%