What is Artificial Analysis Intelligence Index v5?

Artificial Analysis Intelligence Index v5 is a research from Artificial Analysis. Artificial Analysis's aggregate benchmark index that scores language models on intelligence across multiple evaluations.

Who makes Artificial Analysis Intelligence Index v5?

Artificial Analysis Intelligence Index v5 is made by Artificial Analysis.

Does Artificial Analysis Intelligence Index v5 work in the EU?

Available in the EU. The index is published as a public website with no stated geographic restrictions; the terms of use only contain general U.S. export-control language and no EU exclusion. Last verified 14 Nov 2026.

How much does Artificial Analysis Intelligence Index v5 cost?

Free. Leaderboards and index results are viewable on the website without a stated charge; paid offerings, if any, were not listed in the provided material.

Where was Artificial Analysis Intelligence Index v5 shown?

In 2 creator videos by Matt Wolfe, Matthew Berman. Most recently by Matt Wolfe on 9 Nov 2026, at 4:27 in "Anthropic's Fable 5.2 is the World’s Best AI Model".

Artificial Analysis Intelligence Index v5

by Artificial Analysis

Try it Website

Artificial Analysis's aggregate benchmark index that scores language models on intelligence across multiple evaluations.

Available in EUResearch Platforms & InfraResearch

Updated 14 Nov 2026EU status checked 14 Nov 2026Last shown 9 Nov 2026

Artificial Analysis Intelligence Index v5 at a glance

What is Artificial Analysis Intelligence Index v5?: Artificial Analysis Intelligence Index v5 is a research from Artificial Analysis. Artificial Analysis's aggregate benchmark index that scores language models on intelligence across multiple evaluations.
Who makes Artificial Analysis Intelligence Index v5?: Artificial Analysis Intelligence Index v5 is made by Artificial Analysis.
Does Artificial Analysis Intelligence Index v5 work in the EU?: Available in the EU. The index is published as a public website with no stated geographic restrictions; the terms of use only contain general U.S. export-control language and no EU exclusion. Last verified 14 Nov 2026.
How much does Artificial Analysis Intelligence Index v5 cost?: Free. Leaderboards and index results are viewable on the website without a stated charge; paid offerings, if any, were not listed in the provided material.
Where was Artificial Analysis Intelligence Index v5 shown?: In 2 creator videos by Matt Wolfe, Matthew Berman. Most recently by Matt Wolfe on 9 Nov 2026, at 4:27 in “Anthropic's Fable 5.2 is the World’s Best AI Model”.

Sources: the vendor's own pages (linked above) and creator videos at the linked timestamps. EU status is information, not legal advice; see how we check.

About

The Artificial Analysis Intelligence Index is a composite benchmark published by Artificial Analysis, an independent AI benchmarking company. It combines results from a set of evaluations covering areas such as reasoning and knowledge, agentic coding and terminal use, agentic knowledge work, scientific research workflows, long-context reasoning and hallucination rate. Version 5 is the latest iteration referenced by creators, with models ranked by a single aggregate score.

According to a YouTube creator's summary, Gemini 4 Carbon ranks first on this version with a score of 59, ahead of Claude Opus 5.5 at 58 and GPT-6 Astra at 57. The site also reports related measures, including cost per task, time per task, a Coding Agent Index, a Cyber Index and a Decision Index, plus capability breakdowns by domain such as finance, legal, healthcare and engineering. The details of the v5 methodology were not included in the provided page text.

The index is viewed as charts and leaderboards on the Artificial Analysis website, which also offers model and API provider comparisons on speed, latency and price. Results are published by the vendor, so users depend on its testing choices and update schedule.

Key features

Aggregate score across multiple language model evaluations
Covers agentic coding, knowledge work, reasoning and hallucination rate
Companion metrics for cost per task and time per task
Related indexes: Coding Agent, Cyber and Decision
Public leaderboard comparing frontier models

2 moments

As seen on YouTube

Watch the moment: the exact second top creators show it. Videos load only when you press play, via youtube-nocookie.com.

Matt Wolfe9 Nov 2026

Anthropic's Fable 5.2 is the World’s Best AI Model

An aggregate benchmark index that was recently updated. Fable 5.2 tops it, with the highest cost per task at about $26.54 at max reasoning.

4:27Open on YouTube

Matthew Berman29 Oct 2026

Google "Carbon" just won everything...

Newest version of Artificial Analysis's aggregate benchmark index, where Gemini 4 Carbon ranks first with 59 versus Opus 5.5 at 58 and GPT-6 Astra at 57.

1:08Open on YouTube