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.

← Integrity Index

CadEval

CadEval

Rank #8 of 51 · 14 models scored · top score 74

AINTEGRITY 98 / 100

unknown

Strongest on discrimination, weakest on saturation headroom. Caveat: the public test set leaves contamination on the table.

consistent harnesspublic test set

Headroom to ceiling

26 headroomtop 74ceiling 100

Score distribution

median 42o3 7494377

State of the art over time

943777420242025

Integrity breakdown

Weighted composite. Each signal is computed from the source-backed leaderboard below.

Discrimination30%100

Does it still separate models?

Saturation headroom30%95

How far from ceiling / clustered at the top?

Contamination resistance15%100

Public vs held-out; training-leak risk.

Harness comparability15%100

Apples-to-apples, or mixed / vendor-optimized?

Freshness10%96

How old is the benchmark?

Leaderboard

#ModelScoreEvidenceMeasured
1o374unverified2024-12-20
2Gemini 2.5 Pro (Mar 2025)64unverified2025-03-25
3o4-mini62unverified2025-04-16
4o156unverified2024-12-05
5Claude 3.7 Sonnet54unverified2025-02-24
6o3-mini54unverified2025-01-31
7Claude 3.5 Sonnet (October 2024)48unverified2024-10-22
8GPT-4.142unverified2025-04-14
9Gemini 1.5 Pro (Sept 2024)34unverified2024-09-24
10Claude 3.5 Haiku32unverified2024-10-22
11Gemini 2.0 Flash (Feb 2025)30unverified2024-12-11
12GPT-4o (Aug 2024)26unverified2024-05-13
13GPT-4.1 mini16unverified2025-04-14
14Claude 3 Haiku12unverified2024-03-04

At a glance

Models scored
14
Top model
o3
Top score
74
Score spread
62
Headroom
26

Benchmark context

Task type
unknown
Metric
normalized performance in Epoch CSV, rendered as accuracy (%)
Score ceiling
100
Human baseline
unknown
Test-set privacy
unknown
Construction
unknown
Key assumptions
unknown
Limitations
unknown
Contamination history
unknown
Harness notes
Epoch CSV reports model, model version, performance, optimized flag, score source, benchmark release date, and model date.
Authors
unknown
Released
2025-04-22
License
CC BY 4.0
Source
epoch.ai/data/eci_benchmarks.csv

Provenance

14 source records · scores from epoch_benchmarks. Every fact on this page is evidence-linked.

Follow the record

Integrity neighbors

  • ARC-AGI-2A · 99
  • FrontierMath-2025-02-28-PrivateA · 99
  • FrontierMath-Tier-4-v2-PrivateA · 99
  • CursorBenchA · 98
  • FrontierMath-Tiers-1-3-v2-PrivateA · 98
  • GBAEvalA · 98

Top models here

  • o374
  • Gemini 2.5 Pro (Mar 2025)64
  • o4-mini62
  • o156
  • Claude 3.7 Sonnet54

Head-to-head

  • o3 vs Gemini 2.5 Pro (Mar 2025)
  • o3 vs o4-mini
  • Gemini 2.5 Pro (Mar 2025) vs o4-mini