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

Gemini 2.5 Pro (May 2025)

Google DeepMind·released 2025-05-06

independently reproduced

Scored on 9 benchmarks. 22% of its scores are independently reproduced — the rest are self-reported or unverified.

9
Benchmarks scored
0
SOTA leads
95.9
Best score
22%
Reproduced
2 of 9

Benchmark fingerprint

Every score, ranked, colored by evidence status.

  • MATH level 5
    95.9
  • GeoBench
    86
  • Aider polyglot
    76.9
  • Fiction.LiveBench
    66.7
  • GPQA diamond
    55.56
  • DeepResearch Bench
    31.9
  • The Agent Company
    30.3
  • HLE
    13.66
  • VPCT
    10.75

All benchmark scores

BenchmarkScoreEvidence statusMeasured
MATH level 595.9accuracy (%)reproduced· optimized2025-05-06
GeoBench86accuracy (%)unverified2025-05-06
Aider polyglot76.9accuracy (%)unverified· optimized2025-05-06
Fiction.LiveBench66.7accuracy (%)unverified2025-05-06
GPQA diamond55.56accuracy (%)reproduced· optimized2025-05-06
DeepResearch Bench31.9accuracy (%)unverified2025-05-06
The Agent Company30.3accuracy (%)unverified2025-05-06
HLE13.66accuracy (%)unverified2025-05-06
VPCT10.75accuracy (%)unverified2025-05-06
9 source records · epoch_benchmarks. Every score is evidence-linked.

Follow the record

More from Google DeepMind

  • Gemini 3.1 Pro98
  • Gemini 2.5 Pro (Mar 2025)95.56
  • Gemini 3.5 Flash95.55
  • Gemini 3 Flash92.77
  • Gemini 2.5 Pro (Jun 2025)91.7
  • Gemini 3 Pro91.38