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.

← Compare

Claude Opus 4.1 vs GPT-5.2

Verdict

GPT-5.2 leads 11–2 across 13 shared benchmarks.

Claude Opus 4.1 2 · GPT-5.2 11 · higher isn't always better — see caveats.

Per-benchmark head-to-head

Claude Opus 4.1GPT-5.2
  • 49.7
    DeepResearch Bench
    41.12
  • 2.44
    FrontierMath-Tier-4-v2-Private
    31.7
  • 12.63
    FrontierMath-Tiers-1-3-v2-Private
    67.4
  • 43.6
    GDPval
    49.7
  • 69.7
    GPQA diamond
    88.53
  • 7.06
    HLE
    24.16
  • 68.86
    OTIS Mock AIME 2024-2025
    96.11
  • 52
    SimpleBench
    34.96
  • 34.8
    SimpleQA Verified
    38.9
  • 73.35
    SWE-Bench verified
    73.76
  • 38
    Terminal Bench
    64.9
  • 2.5
    VPCT
    76
  • 42.76
    WeirdML
    72.2
A higher number is not always a better model. Each score is task performance under a disclosed harness, and some of these benchmarks are saturated or scored under mixed harnesses. Follow any benchmark to its integrity grade before reading a win as decisive.

Follow the record

These models

  • Claude Opus 4.185.38
  • GPT-5.296.11