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

VPCT

VPCT

Rank #25 of 51 · 26 models scored · top score 86.5

AINTEGRITY 87 / 100

unknown

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

consistent harnesspublic test set

Headroom to ceiling

13.5 headroomtop 86.5ceiling 100

Score distribution

median 10Gemini 3 Pro 86.504590

State of the art over time

0459086.520242025

Integrity breakdown

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

Discrimination30%100

Does it still separate models?

Saturation headroom30%97

How far from ceiling / clustered at the top?

Contamination resistance15%23

Public vs held-out; training-leak risk.

Harness comparability15%100

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

Freshness10%91

How old is the benchmark?

Leaderboard

#ModelScoreEvidenceMeasured
1Gemini 3 Pro86.5unverified2025-11-18
2GPT-5.276unverified2025-12-11
3Gemini 3 Flash58.9unverified2025-12-17
4GPT-549unverified2025-08-07
5GPT-5.138.05unverified2025-11-13
6o4-mini36.25unverified2025-04-16
7o328unverified2024-12-20
8Gemini 2.5 Pro (Mar 2025)22unverified2025-03-25
9Gemini 2.5 Pro (Jun 2025)19.6unverified2025-06-05
10Gemini 2.5 Flash (Sep 2025)19.3unverified2025-09-25
11GPT-4.517.5unverified2025-02-27
12Gemini 2.5 Pro (May 2025)10.75unverified2025-05-06
13GPT-5 mini10.3unverified2025-08-07
14Claude Opus 4.510unverified2025-11-24
15GPT-4o (Nov 2024)10unverified2024-05-13
16Claude Sonnet 4.59.7unverified2025-09-29
17Claude 3.7 Sonnet8.5unverified2025-02-24
18Claude Opus 47unverified2025-05-22
19Gemini 2.5 Flash (Apr 2025)7unverified2025-04-17
20GPT-5 nano5.8unverified2025-08-07
21o15.5unverified2024-12-05
22Claude Opus 4.12.5unverified2025-08-05
23Claude Sonnet 41unverified2025-05-22
24GPT-4o mini1unverified2024-07-18
25Claude 3.5 Sonnet0unverified2024-06-20
26Claude 3.5 Sonnet (October 2024)0unverified2024-10-22

At a glance

Models scored
26
Top model
Gemini 3 Pro
Top score
86.5
Score spread
86.5
Headroom
13.5

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
2025-01-30
License
CC BY 4.0
Source
epoch.ai/data/eci_benchmarks.csv

Provenance

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

Follow the record

Integrity neighbors

  • WeirdMLA · 88
  • Aider polyglotA · 87
  • Fiction.LiveBenchA · 87
  • SimpleBenchA · 86
  • Terminal BenchB · 84
  • SWE-Bench verifiedB · 83

Top models here

  • Gemini 3 Pro86.5
  • GPT-5.276
  • Gemini 3 Flash58.9
  • GPT-549
  • GPT-5.138.05

Head-to-head

  • Gemini 3 Pro vs GPT-5.2
  • Gemini 3 Pro vs Gemini 3 Flash
  • GPT-5.2 vs Gemini 3 Flash