tensor.news
Integrity IndexModelsCompareLabsPapersMethodologyAbout
tensor.news

© 2026

Evidence

  • Integrity Index
  • Model Claim Ledger
  • Compare

Entities

  • Models
  • Benchmarks
  • Labs
  • Papers

Method

  • Methodology
  • About
  • Privacy

Machine

  • Feed (Atom)
  • llms.txt
  • llms-full.txt
  • Sitemap

tensor.news grades benchmarks and tags model claims by evidence status. Every fact carries a source record and provenance. Scores are framed as task performance under a disclosed harness, never as deployed capability.

← Models

Qwen3-235B-A22B

Qwen·released 2025-04-28

independently reproduced

Scored on 7 benchmarks. 29% of its scores are independently reproduced — the rest are self-reported or unverified.

7
Benchmarks scored
0
SOTA leads
83
Best score
29%
Reproduced
2 of 7

Benchmark fingerprint

Every score, ranked, colored by evidence status.

  • Lech Mazur Writing
    83
  • MATH level 5
    68.86
  • Fiction.LiveBench
    67.7
  • GPQA diamond
    60.94
  • Aider polyglot
    59.6
  • WeirdML
    37.28
  • SimpleBench
    17.2

All benchmark scores

BenchmarkScoreEvidence statusMeasured
Lech Mazur Writing83accuracy (%)unverified2025-04-28
MATH level 568.86accuracy (%)reproduced· optimized2025-04-28
Fiction.LiveBench67.7accuracy (%)unverified2025-04-28
GPQA diamond60.94accuracy (%)reproduced· optimized2025-04-28
Aider polyglot59.6accuracy (%)unverified· optimized2025-04-28
WeirdML37.28accuracy (%)unverified2025-04-28
SimpleBench17.2accuracy (%)unverified2025-04-28
7 source records · epoch_benchmarks. Every score is evidence-linked.

Follow the record

More from Qwen

  • Qwen3-Max97.13
  • Qwen3.7-Max94.99
  • Qwen2.5-72B92.67
  • Qwen 3.6 Max (Preview)91.1
  • Qwen2.5-Coder-32B91.1
  • Qwen 3.6 Plus90.55