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 (Apr 2025)

Google DeepMind·released 2025-04-17

independently reproduced

Scored on 10 benchmarks. 10% of its scores are independently reproduced — the rest are self-reported or unverified.

10
Benchmarks scored
0
SOTA leads
76.5
Best score
10%
Reproduced
1 of 10

Benchmark fingerprint

Every score, ranked, colored by evidence status.

  • Lech Mazur Writing
    76.5
  • OTIS Mock AIME 2024-2025
    73.03
  • GeoBench
    73
  • Fiction.LiveBench
    47.2
  • Aider polyglot
    47.1
  • WeirdML
    40.95
  • ARC-AGI
    32.3
  • DeepResearch Bench
    29.19
  • HLE
    7.65
  • VPCT
    7

All benchmark scores

BenchmarkScoreEvidence statusMeasured
Lech Mazur Writing76.5accuracy (%)unverified2025-04-17
OTIS Mock AIME 2024-202573.03accuracy (%)reproduced· optimized2025-04-17
GeoBench73accuracy (%)unverified2025-04-17
Fiction.LiveBench47.2accuracy (%)unverified2025-04-17
Aider polyglot47.1accuracy (%)unverified· optimized2025-04-17
WeirdML40.95accuracy (%)unverified2025-04-17
ARC-AGI32.3accuracy (%)unverified· optimized2025-04-17
DeepResearch Bench29.19accuracy (%)unverified2025-04-17
HLE7.65accuracy (%)unverified2025-04-17
VPCT7accuracy (%)unverified2025-04-17
10 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