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 4.5 vs Kimi K2.5

Verdict

Claude Opus 4.5 leads 12–2 across 15 shared benchmarks.

Claude Opus 4.5 12 · Kimi K2.5 2 · 1 tied · higher isn't always better — see caveats.

Per-benchmark head-to-head

Claude Opus 4.5Kimi K2.5
  • 18.4
    APEX-Agents
    14.4
  • 80
    ARC-AGI
    65.33
  • 37.64
    ARC-AGI-2
    11.81
  • 12
    Chess Puzzles
    12
  • 21.1
    CL-bench
    19.3
  • 81.4
    GPQA diamond
    83.47
  • 21.43
    HLE
    20.56
  • 66.3
    OSWorld
    63.3
  • 86.1
    OTIS Mock AIME 2024-2025
    92.19
  • 17.29
    PostTrainBench
    10.26
  • 54.4
    SimpleBench
    36.16
  • 41.8
    SimpleQA Verified
    33.9
  • 76.65
    SWE-Bench verified
    73.76
  • 63.1
    Terminal Bench
    43.2
  • 63.7
    WeirdML
    45.6
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 4.586.1
  • Kimi K2.592.19