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-V3.2-Exp

DeepSeek·released 2025-09-29

unverified

Scored on 6 benchmarks, leading 1. 0% of its scores are independently reproduced — the rest are self-reported or unverified.

6
Benchmarks scored
1
SOTA leads
83.3
Best score
0%
Reproduced
0 of 6

Leads on: The Agent Company

Benchmark fingerprint

Every score, ranked, colored by evidence status.

  • Fiction.LiveBench
    83.3
  • Aider polyglot
    74.2
  • The Agent Company
    42.9
  • WeirdML
    39.48
  • CL-bench
    13.2
  • CL-bench Life
    9.5

All benchmark scores

BenchmarkScoreEvidence statusMeasured
Fiction.LiveBench83.3accuracy (%)unverified2025-09-29
Aider polyglot74.2accuracy (%)unverified· optimized2025-09-29
The Agent Company42.9accuracy (%)unverified2025-09-29
WeirdML39.48accuracy (%)unverified2025-09-29
CL-bench13.2accuracy (%)unverified2025-09-29
CL-bench Life9.5accuracy (%)unverified2025-09-29
6 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-R193.05
  • DeepSeek-V2 (MoE-236B, May 2024)89.6
  • DeepSeek-V3.287.81

Benchmarks it leads

  • The Agent Company