Multi-task Indic Language Understanding Benchmark (MILU)
We show this table for reference; we do not rank on it.
Culturally grounded knowledge comprehension across ten Indic languages and English.
Benchmark score on MILU — November 13, 2026
We compile the MILU rows from provider self-reports. Claude Fable 5.2 leads the table at 94.1%, followed by Claude Opus 5.5 (93.1%) and Claude Fable 5.1 (93.0%). We do not use these results to rank models overall.
Claude Fable 5.2
Anthropic
Reported by AnthropicClaude Opus 5.5
Anthropic
Reported by AnthropicClaude Fable 5.1
Anthropic
Reported by Anthropic5 modelsMultilingualRefreshingDisplay onlyUpdated November 13, 2026
Benchmark score table (5 models)
All 5 results on this page were reported by the labs that built the models.
| Rank | Model / configuration | Score | Who ran it | Parameters (B) | Open / closed |
|---|---|---|---|---|---|
| 1 | Claude Fable 5.2AnthropicReported by Anthropic | 94.1% | Reported by Anthropic | Not reported | Closed |
| 2 | Claude Opus 5.5AnthropicReported by Anthropic | 93.1% | Reported by Anthropic | Not reported | Closed |
| 3 | Claude Fable 5.1AnthropicReported by Anthropic | 93.0% | Reported by Anthropic | Not reported | Closed |
| 4 | Claude Opus 5AnthropicReported by Anthropic | 92.1% | Reported by Anthropic | Not reported | Closed |
| 5 | Claude Sonnet 5.5AnthropicReported by Anthropic | 91.6% | Reported by Anthropic | Not reported | Closed |
- Lab-reported
- Benchmark-run
- Independent
Among the reported MILU rows, Claude Fable 5.2 is first at 94.1%. The third row is 1.1 points behind. The broader top-10 range is 2.5 points, so many of the published results sit in a relatively narrow band.
5 models have been evaluated on MILU. The benchmark falls in the Multilingual category. MILU is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.
About MILU
Year
2024
Tasks
Knowledge tasks across 11 languages
Format
Average accuracy
Difficulty
Multilingual Indic knowledge
Anthropic reports average accuracy over five max-effort trials without tools or a custom system prompt.
Freshness and provenance
Version
MILU 2024
Refresh cadence
Annual
Staleness state
Refreshing
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 MILU measure?
Culturally grounded knowledge comprehension across ten Indic languages and English.
Which model scores highest on MILU?
Claude Fable 5.2 by Anthropic currently leads with a score of 94.1% on MILU.
How many models are evaluated on MILU?
5 AI models have published results on MILU 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.