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

DeepSeek-R1

DeepSeek·released 2025-01-20

independently reproduced

Scored on 12 benchmarks. 25% of its scores are independently reproduced — the rest are self-reported or unverified.

12
Benchmarks scored
0
SOTA leads
93.05
Best score
25%
Reproduced
3 of 12

Benchmark fingerprint

Every score, ranked, colored by evidence status.

  • MATH level 5
    93.05
  • Lech Mazur Writing
    83
  • Fiction.LiveBench
    69.4
  • GPQA diamond
    62.29
  • Aider polyglot
    56.9
  • OTIS Mock AIME 2024-2025
    53.29
  • WeirdML
    36.49
  • Balrog
    34.9
  • SimpleBench
    17.08
  • ARC-AGI
    15.8
  • ARC-AGI-2
    1.3
  • CritPt
    1.1

All benchmark scores

BenchmarkScoreEvidence statusMeasured
MATH level 593.05accuracy (%)reproduced· optimized2025-01-20
Lech Mazur Writing83accuracy (%)unverified2025-01-20
Fiction.LiveBench69.4accuracy (%)unverified2025-01-20
GPQA diamond62.29accuracy (%)reproduced· optimized2025-01-20
Aider polyglot56.9accuracy (%)unverified· optimized2025-01-20
OTIS Mock AIME 2024-202553.29accuracy (%)reproduced· optimized2025-01-20
WeirdML36.49accuracy (%)unverified2025-01-20
Balrog34.9accuracy (%)unverified2025-01-20
SimpleBench17.08accuracy (%)unverified2025-01-20
ARC-AGI15.8accuracy (%)unverified· optimized2025-01-20
ARC-AGI-21.3accuracy (%)unverified2025-01-20
CritPt1.1accuracy (%)unverified2025-01-20
12 source records · epoch_benchmarks. Every score is evidence-linked.

Follow the record

More from DeepSeek

  • DeepSeek-V4-Pro96.66
  • DeepSeek-R1 (May 2025)96.64
  • DeepSeek-V393.73
  • DeepSeek-V2 (MoE-236B, May 2024)89.6
  • DeepSeek-V3.287.81
  • DeepSeek-V3.185.17