Medical Bench
Agents read longitudinal dermatology charts and write clinical assessments. Expert-authored rubrics grade their findings and management decisions.
Turing reports nine model configurations. Their published scores match the mean clinical task scores to one decimal place. Models cover either nine or 18 tasks.
- Made by
- Turing
- Released
- September 2026
- Kind
- Medicine
- Level
- Expert
- Scored
- Graded by a model
- Answers
- Agent task
- Tasks
- All public
- Measures
- Capability
- Subject
- Clinical medicine
- Language
- English
- Interaction
- Agentic
- Items from
- Real work and use, Written by experts
- Refresh
- One fixed set
- Run by
- A host
- Occupations
- Healthcare practitioners and technical
- In views
- Agents, Biology and medicine
- Strongest model tested
- Claude Opus 5 #11 on the capability index
- Last model added
- August 2026
- Scores refreshed
- November 21st, 2026
Grok 4.6 leads with 88% among models released by August 2026
Best score per model from the benchmark's own leaderboard, read on November 21st, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart.
The newest model with a score here came out in August 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 13 more
- 1 Grok 4.6 xhigh 88%
- 2 Claude Opus 5 max 84.2%
- 3 GPT-5.6 Sol xhigh 69.6%
- 4 Claude Sonnet 5 max 66.5%
- 5 GPT-5.6 Terra max 63.6%
- 6
Gemini 3.7 Flash high 44%
- 7
Gemini 3.1 Pro high 42.4%
- 8
Gemini 3.6 Flash high 32.2%
Source: Official Medical Bench leaderboard
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Grok 4.6 | 88% | xhigh | Aug 2026 | Official leaderboard |
| 2 | Claude Opus 5 | 84.2% | max | Jul 2026 | Official leaderboard |
| 3 | GPT-5.6 Sol | 69.6% | xhigh | Jul 2026 | Official leaderboard |
| 4 | Claude Sonnet 5 | 66.5% | max | Jun 2026 | Official leaderboard |
| 5 | GPT-5.6 Terra | 63.6% | max | Jul 2026 | Official leaderboard |
| 6 | | 44% | high | Aug 2026 | Official leaderboard |
| 7 | | 42.4% | high | Feb 2026 | Official leaderboard |
| 8 | | 32.2% | high | Jul 2026 | Official leaderboard |
The best score, 88%, costs $0.34 a task
Runs on the cost frontier score higher than any run that's cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.
Source: Official Medical Bench leaderboard
| Model | Effort | Harness | Score | Cost per task | Output tokens |
|---|---|---|---|---|---|
| Grok 4.6 | xhigh | 88% | $0.34 | ||
| Claude Opus 5 | max | 84.2% | $1.67 | ||
| GPT-5.6 Sol | xhigh | 69.6% | $1.19 | ||
| Claude Sonnet 5 | max | 66.5% | $0.80 | ||
| GPT-5.6 Terra | max | 63.6% | $0.73 | ||
| Grok 4.6 | max | 63.2% | $0.44 | ||
| high | 44% | $0.42 | |||
| high | 42.4% | $0.22 | |||
| high | 32.2% | $0.39 |
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official Medical Bench leaderboard
Grok 4.6 on the official leaderboard passed 85% in August 2026
Models released before it came out already scored that high. All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official Medical Bench 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 Law, finance and medicine
- MedXpertQA-MM In use Models answer expert-level medical exam questions that combine images with clinical records and test results. Scoring is accuracy.
Qwen 3.8 Max 80.4%
- HealthBench Professional In use 525 physician-authored tasks from real clinician chats (care consult, documentation, research) with physician-adjudicated rubrics. Claude Sonnet 5.5 69.2%
- HealthBench In use 5,000 realistic health conversations graded against physician-written rubrics. A length-adjusted variant penalizes verbosity. GPT-5 67.2%
- FinSearchComp In use Agents search the web to fetch live financial data, look up history, and run multi-step investigations, across 635 expert-written questions. Scored by answer accuracy.
Ling-3.1-flash 81.7%
- KorMedMCQA In use Models answer 7,469 multiple-choice questions from Korean licensing exams for doctors, nurses, pharmacists and dentists, held from 2012 to 2024. o1-preview 92.7%
- Harvey LAB-AA v1.1 In use Artificial Analysis's run of Harvey's Legal Agent Benchmark on Harvey's newer private set of 120 tasks across 24 practice areas. A deliverable with a material hallucination scores zero, and the headline is the share of judges who find every rubric criterion passed. GPT-6.2 Astra 15%
Data audit
All 9 checks pass. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 3 pass 2 notes
- Pass Independent runs Scores from the official leaderboard (8 entries).
- Pass Models scored 8 models we track.
- Pass Newest score The newest is from August 2026, for Gemini 3.7 Flash.
- Note Frontier models None of the top 10 has a score, which is fine for a retired test. The strongest tested is Claude Opus 5 (#11).
- Note Headroom Saturated. The best is 88%, past 85% since August 2026.
Leaderboard import 3 pass
- Pass Importer We read the board on bench.turing.com: 9 rows for models we track at the last read.
- Pass Last read November 21st, 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.