Clinical Benchmarks
Menu

MLCR-AA (Medical Long Context Reasoning)

Questions over long, fragmented medical records; a pass needs a complete answer with nothing that contradicts the record.

Official pageReferences

More about this boardLess

Reasoning over long, fragmented medical records (Wisedocs' MLCR): 60 held-out questions from the expert and compound tiers, each run three times; a response passes when it is judged complete and no claim contradicts the record.

Published by Wisedocs, run by Artificial Analysis, released Aug 2026. 60 private questions over long, fragmented medical records, run 3 times. Pass rate %: responses judged complete with no claim contradicting the record, higher better.

Headline metric
Overall pass rate (expert + compound sets)
Board last updated
5 Nov 2026
Board updated basis
page states no update date; retrieval date
Publisher
Wisedocs, run by Artificial Analysis
Rows
27
Models
27 on the board
Released
Aug 2026
Measured to
Nov 2026
Unit
60 record questions
Documents cited
1

Every result on MLCR-AA, on the board's own scale

27 results, one column each, highest first. Pass rate %: responses judged complete with no claim contradicting the record, higher better.

25 of 27 results
  1. 41.7Claude Sonnet 5.5 (Max, Default Fallback)
  2. 41.7GPT-6 Astra (Max)
  3. 38.3GPT-6.1 Sol (Max)
  4. 38.3Kimi K3.1 Pro
  5. 35.0Claude Haiku 5.5 (Max)
  6. 33.9GPT-6 Sol (Max)
  7. 31.7GPT-5.5 Instant (June 2026)
  8. 30.6GPT-6.1 Terra (Max)
  9. 27.8Claude Opus 5.5
  10. 26.7DeepSeek V4.1 Flash 1017 (Max)
  11. 26.1GPT-6 Luna (Max)
  12. 22.8Qwen3.8 27B (Xhigh)
  13. 22.2Muse Wonder (Max)
  14. 22.2Muse Wonder Full (Max)
  15. 20.0Grok 4.7 (Xhigh)
  16. 20.0Qwen3.8 Max (0902)
  17. 17.8Beam 1.0
  18. 17.2GLM-5.3 (Max)
  19. 16.7MiniMax M3.1 Flash
  20. 15.6Claude Fable 5.1 (Max, Default Fallback)
  21. 15.0Ling 3.1 Flash 1019
  22. 13.3Nemotron 3.5 Super VL
  23. 12.8MiMo V2.6 Pro
  24. 12.2K2 Horizon 375B A23B
  25. 10.6Gemini 3.8 Flash (High)

Column colour is the model's lab. The table below says who ran each result.

Every published result, ranked on this board alone

27 rows. Higher is better. Bars run from 0 to 100; whiskers show a published interval. Select a row for its quote, page and document.

1Claude Sonnet 5.5 (Max, Default Fallback)Anthropic · overall pass rate41.7Independent runNov 2026[1]
Claude Sonnet 5.5 (Max, Default Fallback): 41.7% (Score: share of responses judged complete with no claim contradicting the record)
As printed
41.7%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate, Max, Default Fallback
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
1GPT-6 Astra (Max)OpenAI · overall pass rate41.7Independent runNov 2026[1]
GPT-6 Astra (Max): 41.7% (Score: share of responses judged complete with no claim contradicting the record)
As printed
41.7%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate, Max
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
3GPT-6.1 Sol (Max)OpenAI · overall pass rate38.3Independent runNov 2026[1]
GPT-6.1 Sol (Max): 38.3% (Score: share of responses judged complete with no claim contradicting the record)
As printed
38.3%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate, Max
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
3Kimi K3.1 ProMoonshot AI · overall pass rate38.3Independent runNov 2026[1]
Kimi K3.1 Pro: 38.3% (Score: share of responses judged complete with no claim contradicting the record)
As printed
38.3%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
5Claude Haiku 5.5 (Max)Anthropic · overall pass rate35.0Independent runNov 2026[1]
Claude Haiku 5.5 (Max): 35.0% (Score: share of responses judged complete with no claim contradicting the record)
As printed
35.0%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate, Max
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
6GPT-6 Sol (Max)OpenAI · overall pass rate33.9Independent runNov 2026[1]
GPT-6 Sol (Max): 33.9% (Score: share of responses judged complete with no claim contradicting the record)
As printed
33.9%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate, Max
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
7GPT-5.5 Instant (June 2026)OpenAI · overall pass rate31.7Independent runNov 2026[1]
GPT-5.5 Instant (June 2026): 31.7% (Score: share of responses judged complete with no claim contradicting the record)
As printed
31.7%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate, June 2026
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
8GPT-6.1 Terra (Max)OpenAI · overall pass rate30.6Independent runNov 2026[1]
GPT-6.1 Terra (Max): 30.6% (Score: share of responses judged complete with no claim contradicting the record)
As printed
30.6%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate, Max
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
9Claude Opus 5.5Anthropic · overall pass rate27.8Independent runNov 2026[1]
Claude Opus 5.5: 27.8% (Score: share of responses judged complete with no claim contradicting the record)
As printed
27.8%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
10DeepSeek V4.1 Flash 1017 (Max)DeepSeek · overall pass rate26.7Independent runNov 2026[1]
DeepSeek V4.1 Flash 1017 (Max): 26.7% (Score: share of responses judged complete with no claim contradicting the record)
As printed
26.7%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate, Max
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
11GPT-6 Luna (Max)OpenAI · overall pass rate26.1Independent runNov 2026[1]
GPT-6 Luna (Max): 26.1% (Score: share of responses judged complete with no claim contradicting the record)
As printed
26.1%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate, Max
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
12Qwen3.8 27B (Xhigh)Alibaba · overall pass rate22.8Independent runNov 2026[1]
Qwen3.8 27B (Xhigh): 22.8% (Score: share of responses judged complete with no claim contradicting the record)
As printed
22.8%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate, Xhigh
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
13Muse Wonder (Max)Meta · overall pass rate22.2Independent runNov 2026[1]
Muse Wonder (Max): 22.2% (Score: share of responses judged complete with no claim contradicting the record)
As printed
22.2%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate, Max
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
13Muse Wonder Full (Max)Meta · overall pass rate22.2Independent runNov 2026[1]
Muse Wonder Full (Max): 22.2% (Score: share of responses judged complete with no claim contradicting the record)
As printed
22.2%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate, Max
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
15Grok 4.7 (Xhigh)SpaceX AI · overall pass rate20.0Independent runNov 2026[1]
Grok 4.7 (Xhigh): 20.0% (Score: share of responses judged complete with no claim contradicting the record)
As printed
20.0%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate, Xhigh
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
15Qwen3.8 Max (0902)Alibaba · overall pass rate20.0Independent runNov 2026[1]
Qwen3.8 Max (0902): 20.0% (Score: share of responses judged complete with no claim contradicting the record)
As printed
20.0%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate, 0902
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
17Beam 1.0Reflection AI · overall pass rate17.8Independent runNov 2026[1]
Beam 1.0: 17.8% (Score: share of responses judged complete with no claim contradicting the record)
As printed
17.8%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
18GLM-5.3 (Max)Zhipu · overall pass rate17.2Independent runNov 2026[1]
GLM-5.3 (Max): 17.2% (Score: share of responses judged complete with no claim contradicting the record)
As printed
17.2%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate, Max
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
19MiniMax M3.1 FlashMiniMax · overall pass rate16.7Independent runNov 2026[1]
MiniMax-M3.1-Flash: 16.7% (Score: share of responses judged complete with no claim contradicting the record)
As printed
16.7%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
20Claude Fable 5.1 (Max, Default Fallback)Anthropic · overall pass rate15.6Independent runNov 2026[1]
Claude Fable 5.1 (Max, Default Fallback): 15.6% (Score: share of responses judged complete with no claim contradicting the record)
As printed
15.6%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate, Max, Default Fallback
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
21Ling 3.1 Flash 1019Ant Group · overall pass rate15.0Independent runNov 2026[1]
Ling 3.1 Flash 1019: 15.0% (Score: share of responses judged complete with no claim contradicting the record)
As printed
15.0%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
22Nemotron 3.5 Super VLNVIDIA · overall pass rate13.3Independent runNov 2026[1]
Nemotron 3.5 Super VL: 13.3% (Score: share of responses judged complete with no claim contradicting the record)
As printed
13.3%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
23MiMo V2.6 ProXiaomi · overall pass rate12.8Independent runNov 2026[1]
MiMo-V2.6-Pro: 12.8% (Score: share of responses judged complete with no claim contradicting the record)
As printed
12.8%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
24K2 Horizon 375B A23BInstitute of Foundation Models · overall pass rate12.2Independent runNov 2026[1]
K2 Horizon 375B A23B: 12.2% (Score: share of responses judged complete with no claim contradicting the record)
As printed
12.2%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
25Gemini 3.8 Flash (High)Google · overall pass rate10.6Independent runNov 2026[1]
Gemini 3.8 Flash (High): 10.6% (Score: share of responses judged complete with no claim contradicting the record)
As printed
10.6%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate, High
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
26Muse Spark Open (Xhigh)Meta · overall pass rate7.8Independent runNov 2026[1]
Muse Spark Open (Xhigh): 7.8% (Score: share of responses judged complete with no claim contradicting the record)
As printed
7.8%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate, Xhigh
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
27Inkling (Xhigh)Thinking Machines · overall pass rate6.1Independent runNov 2026[1]
Inkling (Xhigh): 6.1% (Score: share of responses judged complete with no claim contradicting the record)
As printed
6.1%
Where
artificialanalysis.ai/evaluations/mlcr-aa, Score chart (mlcrAaOverallPassRate)
Setting
overall pass rate, Xhigh
Reported by
independent run
Document
[1] MLCR-AA (Medical Long Context Reasoning), Artificial Analysis
References

1 document supports this board

  1. 1.MLCR-AA (Medical Long Context Reasoning), Artificial Analysisofficial leaderboard, Artificial Analysis, 5 Nov 2026. Every result it supports