Integrity Index Models Compare Labs Papers Methodology About

© 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.

DeepSeek

26 models tracked · known for DeepSeek-R1, DeepSeek-V4-Flash, DeepSeek-V3.2

26

Models tracked

26

In our data

96.66

Best score

3

Notable

Models

DeepSeek-V4-Proindependently reproduced96.66 DeepSeek-R1 (May 2025)independently reproduced96.64 DeepSeek-V3independently reproduced93.73 DeepSeek-R1independently reproduced93.05 DeepSeek-V2 (MoE-236B, May 2024)self-reported89.6 DeepSeek-V3.2independently reproduced87.81 DeepSeek-V3.1unverified85.17 DeepSeek-V3.2-Expunverified83.3 DeepSeek-V3 (Mar 2025)independently reproduced77 DeepSeek-Coder-V2-Lite-Baseself-reported67.1 DeepSeek Coder 33Bself-reported35.4 DeepSeek Coder 6.7Bself-reported21.3 DeepSeek Coder 1.3Bself-reported6.6 DeepSeek-R1unverified—DeepSeek-V4-Flashunverified—DeepSeek-V3.2unverified—DeepSeek-R1-0528unverified—DeepSeek-R1-0528-Qwen3-8Bunverified—DeepSeek-V4-Prounverified—DeepSeek-V2-Lite-Chatunverified—DeepSeek-Coder-V2-Lite-Instructunverified—DeepSeek-V3unverified—DeepSeek-V3-0324unverified—DeepSeek-R1-Distill-Qwen-32Bunverified—DeepSeek-R1-Distill-Qwen-1.5Bunverified—DeepSeek-R1-Distill-Llama-70Bunverified—

116 source records · huggingface_models, epoch_benchmarks. Every score is evidence-linked.

Follow the record

Other labs

Qwen
Meta AI
OpenAI
NVIDIA
Google DeepMind
EleutherAI
Zhipu AI
Hugging Face

Explore the record

Benchmark Integrity Index
Model Claim Ledger
Head-to-head comparisons