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 Flash (May 2025)

Google DeepMind·released 2025-05-20

independently reproduced

Scored on 8 benchmarks. 13% of its scores are independently reproduced — the rest are self-reported or unverified.

8
Benchmarks scored
0
SOTA leads
77.8
Best score
13%
Reproduced
1 of 8

Benchmark fingerprint

Every score, ranked, colored by evidence status.

  • Fiction.LiveBench
    77.8
  • GeoBench
    76
  • OTIS Mock AIME 2024-2025
    70.8
  • Aider polyglot
    55.1
  • WeirdML
    40.95
  • ARC-AGI
    33.3
  • HLE
    6.47
  • ARC-AGI-2
    2.54

All benchmark scores

BenchmarkScoreEvidence statusMeasured
Fiction.LiveBench77.8accuracy (%)unverified2025-05-20
GeoBench76accuracy (%)unverified2025-05-20
OTIS Mock AIME 2024-202570.8accuracy (%)reproduced· optimized2025-05-20
Aider polyglot55.1accuracy (%)unverified· optimized2025-05-20
WeirdML40.95accuracy (%)unverified2025-05-20
ARC-AGI33.3accuracy (%)unverified· optimized2025-05-20
HLE6.47accuracy (%)unverified2025-05-20
ARC-AGI-22.54accuracy (%)unverified2025-05-20
8 source records · epoch_benchmarks. Every score is evidence-linked.

Follow the record

More from Google DeepMind

  • Gemini 3.1 Pro98
  • Gemini 2.5 Pro (May 2025)95.9
  • 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