Skip to main content

PolyMath

A multilingual mathematical reasoning benchmark that tests whether math performance transfers across languages rather than only in English.

How BenchLM shows PolyMath right now

BenchLM is tracking PolyMath in the local dataset, but exact-source verification records for these rows are still being attached. To avoid a blank benchmark page, BenchLM shows the current tracked rows below as a display-only reference table.

These tracked rows are useful for inspection and spot-checking, but until exact-source attachments are completed they should not be treated as fully verified public benchmark rows.

5 tracked modelsLocal tracked rowsAwaiting exact-source attachmentsDisplay only

Tracked score on PolyMath — April 16, 2026

BenchLM mirrors the published tracked score view for PolyMath. Claude Opus 4.5 leads the public snapshot at 79.0% , followed by Qwen3.6 Plus (77.4%) and Qwen3.5 397B (73.3%). BenchLM does not use these results to rank models overall.

5 modelsMultilingualCurrentDisplay onlyUpdated April 16, 2026

The published PolyMath snapshot is tightly clustered at the top: Claude Opus 4.5 sits at 79.0%, while the third row is only 5.7 points behind. The broader top-10 spread is 35.9 points, so the benchmark still separates strong models even when the leaders cluster.

5 models have been evaluated on PolyMath. The benchmark falls in the Multilingual category. This category carries a 7% weight in BenchLM.ai's overall scoring system. PolyMath is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About PolyMath

Year

2026

Tasks

Multilingual math problems

Format

Cross-lingual mathematical reasoning

Difficulty

Advanced multilingual reasoning

PolyMath isolates cross-lingual math transfer rather than general chat quality. It is useful for spotting models that keep surface fluency in other languages but lose structured reasoning quality.

BenchLM freshness & provenance

Version

PolyMath 2026

Refresh cadence

Quarterly

Staleness state

Current

Question availability

Public benchmark set

CurrentDisplay only

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.

Tracked score table (5 models)

1
Claude Opus 4.5claude-opus-4-5
79.0%
2
Qwen3.6 Plusqwen3-6-plus
77.4%
3
Qwen3.5 397Bqwen3-5-397b
73.3%
4
GLM-5glm-5
65.2%
5
Kimi K2.5kimi-k2-5
43.1%

FAQ

What does PolyMath measure?

A multilingual mathematical reasoning benchmark that tests whether math performance transfers across languages rather than only in English.

Which model leads the published PolyMath snapshot?

Claude Opus 4.5 currently leads the published PolyMath snapshot with a tracked score of 79.0%. BenchLM shows this benchmark for display only and does not use it in overall rankings.

How many models are evaluated on PolyMath?

5 AI models are included in BenchLM's mirrored PolyMath snapshot, based on the public leaderboard captured on April 16, 2026.

Last updated: April 16, 2026 · mirrored from the public benchmark leaderboard

The AI models change fast. We track them for you.

For engineers, researchers, and the plain curious — a weekly brief on new models, ranking shifts, and pricing changes.

Free. No spam. Unsubscribe anytime.