ChemDIRT
Models answer chemistry questions while the evaluator varies instructions and molecular representations. The authors measure accuracy and consistency across variants.
- Made by
- Eric Inae and colleagues
- Released
- March 2026
- Kind
- Science questions
- Level
- University
- Scored
- Checked automatically
- Answers
- Short answer
- Tasks
- All public
- Measures
- Capability
- Subject
- Chemical sciences
- Language
- English
- In and out
- Text, Molecules and sequences → Text
- Interaction
- Single turn
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Chemistry and materials
- Strongest model tested
- Claude Sonnet 5 #35 on the capability index
- Last model added
- July 2026
- Scores refreshed
- November 20th, 2026
ChemDFM-R tops the leaderboard with 77.7% Claude Sonnet 5 leads with 48.7% among models released by July 2026
Best score per entry from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. The paper gives no overall score, so the headline uses its first measure, exact match for IUPAC-to-SMILES translation. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is Claude Sonnet 5, at 48.7%. This view shows only models we track, each at its best.
The newest model with a score here came out in July 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6 Astra
- and 16 more
- 77.7 ChemDFM-R
- 74.0 ChemDFM-v2
- 72.8 ChemR
- 48.7 Claude Sonnet 5
- 25.0 ChemDFM
- 17.5 ChemVLM-8B
- 15.8 GPT-5.6 Luna
- 14.5
Gemini 3.5 Flash-Lite
- 11.7 Kale-Chem
- 2.20 ChemLLM
- 1.80 RetroDFM-R
- 1.00 ChemM-R1
- 0.0 Falcon
- 0.0 Galactica
- 0.0 Llama3
- 0.0 Mistral
- 0.0
Llama 2-7B
- 0.0
Qwen2-7B-Instruct
- 0.0
Gemma 2 9B
- 0.0
Gemma 3 4B
- 0.0
Qwen3-8B
- 1 ChemDFM-R 77.7%
- 2 ChemDFM-v2 74%
- 3 ChemR 72.8%
- 4 Claude Sonnet 5 48.7%
- 5 ChemDFM 25%
- 6 ChemVLM-8B 17.5%
- 7 GPT-5.6 Luna 15.8%
- 8
Gemini 3.5 Flash-Lite 14.5%
- 9 Kale-Chem 11.7%
- 10 ChemLLM 2.20%
- 11 RetroDFM-R 1.80%
- 12 ChemM-R1 1.00%
- 13 Falcon 0%
- 14 Galactica 0%
- 15 Llama3 0%
- 16 Mistral 0%
- 17
Llama 2-7B 0%
- 18
Qwen2-7B-Instruct 0%
- 19
Gemma 2 9B 0%
- 20
Gemma 3 4B 0%
- 21
Qwen3-8B 0%
Source: The ChemDIRT paper
- 1 Claude Sonnet 5 48.7%
- 2 GPT-5.6 Luna 15.8%
- 3
Gemini 3.5 Flash-Lite 14.5%
- 4
Llama 2-7B 0%
- 5
Qwen2-7B-Instruct 0%
- 6
Gemma 2 9B 0%
- 7
Gemma 3 4B 0%
- 8
Qwen3-8B 0%
Source: The ChemDIRT paper
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | ChemDFM-R | 77.7% | Official leaderboard | ||
| 2 | ChemDFM-v2 | 74% | Official leaderboard | ||
| 3 | ChemR | 72.8% | Official leaderboard | ||
| 4 | Claude Sonnet 5 | 48.7% | Jun 2026 | Official leaderboard | |
| 5 | ChemDFM | 25% | Official leaderboard | ||
| 6 | ChemVLM-8B | 17.5% | Official leaderboard | ||
| 7 | GPT-5.6 Luna | 15.8% | Jul 2026 | Official leaderboard | |
| 8 | | 14.5% | Jul 2026 | Official leaderboard | |
| 9 | Kale-Chem | 11.7% | Official leaderboard | ||
| 10 | ChemLLM | 2.20% | Official leaderboard | ||
| 11 | RetroDFM-R | 1.80% | Official leaderboard | ||
| 12 | ChemM-R1 | 1.00% | Official leaderboard | ||
| 13 | Falcon | 0% | Official leaderboard | ||
| 14 | Galactica | 0% | Official leaderboard | ||
| 15 | Llama3 | 0% | Official leaderboard | ||
| 16 | Mistral | 0% | Official leaderboard | ||
| 17 | | 0% | Jul 2023 | Official leaderboard | |
| 18 | | 0% | Jun 2024 | Official leaderboard | |
| 19 | | 0% | Jun 2024 | Official leaderboard | |
| 20 | | 0% | Mar 2025 | Official leaderboard | |
| 21 | | 0% | Apr 2025 | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Claude Sonnet 5 | 48.7% | Jun 2026 | Official leaderboard | |
| 2 | GPT-5.6 Luna | 15.8% | Jul 2026 | Official leaderboard | |
| 3 | | 14.5% | Jul 2026 | Official leaderboard | |
| 4 | | 0% | Jul 2023 | Official leaderboard | |
| 5 | | 0% | Jun 2024 | Official leaderboard | |
| 6 | | 0% | Jun 2024 | Official leaderboard | |
| 7 | | 0% | Mar 2025 | Official leaderboard | |
| 8 | | 0% | Apr 2025 | Official leaderboard |
2 models have held number one in the benchmark’s paper since its release in March 2026
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: The ChemDIRT paper
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official ChemDIRT leaderboard
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Science and reasoning
- GPQA Diamond In use 198 graduate-level 'Google-proof' multiple-choice questions in biology, chemistry and physics.
Gemini 4 Carbon 97%
- Terminal-Bench-Science In use 70 scientific research-workflow tasks across five science areas, run in Harbor containers. Hidden tests give the resolution rate. GPT-6 Astra 68.1%
- BioMysteryBench In use Bioinformatics problems on anonymized real datasets that require real analysis with tools. Rubric-graded, with human-solvable and human-difficult splits. Claude Fable 5.2 83.7%
- FrontierScience In use Models solve physics, chemistry, and biology problems in an Olympiad track graded by short answer and a Research track graded by 10-point rubrics.
Intern-S2-397B 87%
- HiPhO In use Models solve theory problems from 2024 and 2025 physics olympiad exams, some with diagrams. Official marking schemes grade answers and steps, mapped to medal levels.
Gemini 3 Pro 78.3%
- LABBench2 In use 11 biology research sub-tasks (literature, figures, tables, patents, trials and more) with tool access. Macro-averaged accuracy. GPT-5.2 60.9%
Data audit
All 10 checks pass. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
- Pass Independent runs Scores from the benchmark’s paper (21 entries).
- Pass Models scored 8 models we track, plus 13 other entries on its leaderboard.
- Pass Newest score The newest is from July 2026, for Gemini 3.5 Flash-Lite.
- Note Frontier models None of the top 10 has a score. The strongest tested is Claude Sonnet 5 (#35).
- Pass Headroom The best is 48.7%, under 85%, so it still separates models.
Leaderboard import 3 pass
- Pass Importer We read the board on arxiv.org: 8 rows for models we track and 13 other entries at the last read.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.