Medical Long Context Reasoning (MLCR-AA) (MLCR-AA)
We show this table for reference; we do not rank on it.
An open benchmark from Wisedocs, run by Artificial Analysis, measuring how well models reason over long, fragmented medical records with multi-document synthesis.
Benchmark score on MLCR-AA — November 13, 2026
We compile the MLCR-AA rows from secondary reports. Claude Fable 5.2 leads the table at 43.9%, followed by GPT-6 Astra (41.7%) and Claude Sonnet 5.5 (41.7%). We do not use these results to rank models overall.
Claude Fable 5.2
Anthropic
Carried from an upstream pageGPT-6 Astra
OpenAI
Carried from an upstream pageClaude Sonnet 5.5
Anthropic
Carried from an upstream page21 modelsReasoningCurrentDisplay onlyUpdated November 13, 2026
Benchmark score table (21 models)
None of the 21 results on this page has a documented run by the benchmark or an independent evaluator: 21 carried from an upstream page, where who ran them is not recorded.
| Rank | Model / configuration | Score | Who ran it | Parameters (B) | Open / closed |
|---|---|---|---|---|---|
| 1 | Claude Fable 5.2AnthropicCarried from an upstream page | 43.9% | Carried from an upstream page | Not reported | Closed |
| 2 | GPT-6 AstraOpenAICarried from an upstream page | 41.7% | Carried from an upstream page | Not reported | Closed |
| 3 | Claude Sonnet 5.5AnthropicCarried from an upstream page | 41.7% | Carried from an upstream page | Not reported | Closed |
| 4 | Kimi K3.1 ProMoonshot AICarried from an upstream page | 38.3% | Carried from an upstream page | Not reported | Open |
| 5 | GPT-6.1 SolOpenAICarried from an upstream page | 38.3% | Carried from an upstream page | Not reported | Closed |
| 6 | Claude Haiku 5.5AnthropicCarried from an upstream page | 35.0% | Carried from an upstream page | Not reported | Closed |
| 7 | GPT-6.1 TerraOpenAICarried from an upstream page | 30.6% | Carried from an upstream page | Not reported | Closed |
| 8 | Claude Opus 5.5AnthropicCarried from an upstream page | 27.8% | Carried from an upstream page | Not reported | Closed |
| 9 | GPT-6 LunaOpenAICarried from an upstream page | 26.1% | Carried from an upstream page | Not reported | Closed |
| 10 | Qwen3.8-27BAlibabaCarried from an upstream page | 22.8% | Carried from an upstream page | Not reported | Open |
| 11 | Muse Wonder FullMetaCarried from an upstream page | 22.2% | Carried from an upstream page | Not reported | Closed |
| 12 | Muse WonderMetaCarried from an upstream page | 22.2% | Carried from an upstream page | Not reported | Closed |
| 13 | Inkling 2 SmallThinking Machines LabCarried from an upstream page | 20.6% | Carried from an upstream page | Not reported | Closed |
| 14 | Grok 4.7xAICarried from an upstream page | 20.0% | Carried from an upstream page | Not reported | Closed |
| 15 | Gemini 4 CarbonGoogleCarried from an upstream page | 19.4% | Carried from an upstream page | Not reported | Closed |
| 16 | Muse EmberMetaCarried from an upstream page | 18.3% | Carried from an upstream page | Not reported | Closed |
| 17 | Beam 1.0 501B-A23BReflection AICarried from an upstream page | 17.8% | Carried from an upstream page | Not reported | Open |
| 18 | Mistral Large 4MistralCarried from an upstream page | 17.8% | Carried from an upstream page | Not reported | Open |
| 19 | GLM-5.3Z.AICarried from an upstream page | 17.2% | Carried from an upstream page | Not reported | Open |
| 20 | Claude Fable 5.1AnthropicCarried from an upstream page | 15.6% | Carried from an upstream page | Not reported | Closed |
| 21 | MiMo-V2.6-ProXiaomiCarried from an upstream page | 12.8% | Carried from an upstream page | Not reported | Open |
- Lab-reported
- Benchmark-run
- Independent
Among the reported MLCR-AA rows, Claude Fable 5.2 is first at 43.9%. The third row is 2.2 points behind. The broader top-10 range is 21.1 points, so the table still separates the published systems.
21 models have been evaluated on MLCR-AA. The benchmark falls in the Reasoning category. MLCR-AA is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.
About MLCR-AA
Year
2026
Tasks
Long, fragmented medical-record reasoning
Format
Accuracy
Difficulty
Long-context medical reasoning
Independently benchmarked by Artificial Analysis. Display-only; not yet admitted as independent-run evidence.
Freshness and provenance
Version
MLCR-AA 2026
Refresh cadence
Quarterly
Staleness state
Current
Question availability
Public benchmark set
BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.
Questions
What does MLCR-AA measure?
An open benchmark from Wisedocs, run by Artificial Analysis, measuring how well models reason over long, fragmented medical records with multi-document synthesis.
Which model scores highest on MLCR-AA?
Claude Fable 5.2 by Anthropic currently leads with a score of 43.9% on MLCR-AA.
How many models are evaluated on MLCR-AA?
21 AI models have published results on MLCR-AA in the BenchLM catalog.
Know when it’s worth switching models
The model to choose, the cheaper alternative, and the release we would wait on.
Read a sample issueJoin 7,000+ subscribers.
One email each week. Unsubscribe anytime.