Benchmark Almanac

Hx2Dx ARENA

GPT-6 Luna 83.9% leads since November 2026
Not in model cards

Models interview a simulated patient, request examinations and tests, and give a diagnosis. A judge compares the diagnosis with the published case.

Released
November 2026
Kind
Medicine
Level
Expert
Scored
Graded by a model
Answers
Agent task
Measures
Capability
Language
English
In and out
Text, Images → Text, Actions
Interaction
Agentic
Environment
A simulator
Refresh
One fixed set
Run by
A host
Occupations
Healthcare practitioners and technical
Strongest model tested
DeepSeek V4.1 Flash #34 on the capability index
Last model added
September 2026
Scores refreshed
November 20th, 2026
7
Models scored in its paper

GPT-6 Luna leads with 83.9% among models released by September 2026

Best score per model from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. 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 September 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 GPT-6 Luna Static 83.9%
  2. 2 GLM-5.3-Flash Static 79.7%
  3. 3 DeepSeek V4.1 Flash Static 75.1%
  4. 4 Gemini 3.5 Flash-Lite Static 74.7%
  5. 5 Llama 4 Scout Static 57.9%
  6. 6 Ministral 8B Static 56.1%
  7. 7 Gemma 3 27B Static 55.1%

Source: The Hx2Dx ARENA paper

#ModelScoreSettingDateSource
1 GPT-6 Luna 83.9% Static Sep 2026 Official leaderboard
2 GLM-5.3-Flash 79.7% Static Aug 2026 Official leaderboard
3 DeepSeek V4.1 Flash 75.1% Static Sep 2026 Official leaderboard
4 Gemini 3.5 Flash-Lite 74.7% Static Jul 2026 Official leaderboard
5 Llama 4 Scout 57.9% Static Apr 2025 Official leaderboard
6 Ministral 8B 56.1% Static Oct 2024 Official leaderboard
7 Gemma 3 27B 55.1% Static Mar 2025 Official leaderboard

Scores at each effort level

Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.

Lowest effort to highest, light to dark
GPT-6 Luna
Conversatio… Static
83.9%
GLM-5.3-Flash
Conversatio… Static
79.7%
DeepSeek V4.1 Flash
Conversatio… Static
75.1%
Gemini 3.5 Flash-Lite
Conversatio… Static
74.7%
Llama 4 Scout
Conversatio… Static
57.9%
Ministral 8B
Conversatio… Static
56.1%
Gemma 3 27B
Conversatio… Static
55.1%

Source: The Hx2Dx ARENA paper

ModelEffortHarnessScoreCost per taskOutput tokens
GPT-6 Luna Static 83.9%
GLM-5.3-Flash Static 79.7%
DeepSeek V4.1 Flash Static 75.1%
Gemini 3.5 Flash-Lite Static 74.7%
Llama 4 Scout Static 57.9%
Ministral 8B Static 56.1%
Gemma 3 27B Static 55.1%
GLM-5.3-Flash Conversational 54.4%
DeepSeek V4.1 Flash Conversational 50.3%
GPT-6 Luna Conversational 47.4%
Gemini 3.5 Flash-Lite Conversational 38.5%
Gemma 3 27B Conversational 25.7%
Llama 4 Scout Conversational 23.6%
Ministral 8B Conversational 21.9%

Number one over time

Lab claims count from the day the card came out, independent runs from the model’s release.

OpenAI
Benchmark paper 1 leader
GPT-6 Luna 83.9%

Source: The Hx2Dx ARENA paper

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
In the benchmark’s paper Best so far
0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) GPT-6 Luna GLM-5.3-Flash Gemini 3.5 Flash-Lite Llama 4 Scout Ministral 8B 0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) GPT-6 Luna Gemini 3.5 Flash-Lite

Source: Official Hx2Dx ARENA 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

All 637 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

9 of 10 checks pass, 1 warning. Look first at the Task access check. We don’t know whether the tasks are public. Every benchmark

Catalog record 1 warning 1 pass
  • Warning Task access We don’t know whether the tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
  • Pass Independent runs Scores from the benchmark’s paper (7 entries).
  • Pass Models scored 7 models we track.
  • Pass Newest score The newest is from September 2026, for GPT-6 Luna.
  • Note Frontier models None of the top 10 has a score. The strongest tested is DeepSeek V4.1 Flash (#34).
  • Pass Headroom The best is 83.9%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on arxiv.org: 14 rows for models we track 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.

How we check