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

GLM-5.2

Zhipu AI·released 2026-06-16

independently reproduced

Scored on 15 benchmarks, leading 1. 47% of its scores are independently reproduced — the rest are self-reported or unverified.

15
Benchmarks scored
1
SOTA leads
89.14
Best score
47%
Reproduced
7 of 15

Leads on: PostTrainBench

Benchmark fingerprint

Every score, ranked, colored by evidence status.

  • GPQA diamond
    89.14
  • OTIS Mock AIME 2024-2025
    86.38
  • SWE-Bench verified
    78.7
  • ARC-AGI
    77
  • WeirdML
    70.12
  • FrontierMath-Tiers-1-3-v2-Private
    59.21
  • CursorBench
    54.6
  • SimpleQA Verified
    38.1
  • APEX-Agents
    35.6
  • PostTrainBench
    34.29
  • FrontierMath-Tier-4-v2-Private
    29.27
  • ARC-AGI-2
    22.78
  • Chess Puzzles
    21
  • CritPt
    20.86
  • GBAEval
    0.02

All benchmark scores

BenchmarkScoreEvidence statusMeasured
GPQA diamond89.14accuracy (%)reproduced· optimized2026-06-16
OTIS Mock AIME 2024-202586.38accuracy (%)reproduced· optimized2026-06-16
SWE-Bench verified78.7accuracy (%)reproduced· optimized2026-06-16
ARC-AGI77accuracy (%)unverified· optimized2026-06-16
WeirdML70.12accuracy (%)unverified2026-06-16
FrontierMath-Tiers-1-3-v2-Private59.21accuracy (%)reproduced2026-06-16
CursorBench54.6accuracy (%)unverified2026-06-16
SimpleQA Verified38.1accuracy (%)reproduced2026-06-16
APEX-Agents35.6accuracy (%)unverified2026-06-16
PostTrainBench34.29accuracy (%)unverified2026-06-16
FrontierMath-Tier-4-v2-Private29.27accuracy (%)reproduced2026-06-16
ARC-AGI-222.78accuracy (%)unverified2026-06-16
Chess Puzzles21accuracy (%)reproduced2026-06-16
CritPt20.86accuracy (%)unverified2026-06-16
GBAEval0.02accuracy (%)unverified2026-06-16
15 source records · epoch_benchmarks. Every score is evidence-linked.

Follow the record

More from Zhipu AI

  • GLM-5.192.21
  • GLM-583.76
  • GLM-4.783.32
  • chatglm2-6b54.3
  • GLM-4.624.5
  • GLM-4.7-Flash

Head-to-head

  • vs PowerMoE-3b
  • vs Claude Fable 5
  • vs GPT-5.5
  • vs Gemini 3.1 Pro
  • vs Llama 3.1-405B
  • vs phi-3-small 7.4B

Benchmarks it leads

  • PostTrainBench