Independent evaluation platform
Artificial Analysis
A broad view of model capability, with the evaluation recipe made explicit.
Key information
- Focus
- Language-model capability across a changing evaluation suite.
- Measure
- Artificial Analysis Intelligence Index; separate capability and performance measures.
- Version / scope
- Intelligence Index v4.3.2
- Update pattern
- Models and methodology are updated as the publisher releases evaluations; no fixed schedule is promised here.
- Publisher update
- Not consistently disclosed
- Editorial verification
- 15 October 2026
In plain English
What makes it different?
Artificial Analysis runs evaluations across technical tasks and combines selected results into its Intelligence Index. It also reports other measures separately. The index offers a useful starting point when choosing what to investigate, but its definition is narrower than everything an AI system can do.
Sources and attribution are listed below.
What it measures
Language-model capability across a changing evaluation suite.
Strengths
- One documented framework spans several kinds of capability.
- Component-level methods help explain what sits behind a headline number.
Limitations
- An aggregate reflects the publisher's task selection and weights, rather than a universal definition of intelligence.
- Some inputs and test environments are private. A website listing is not a licence to reproduce its results.
How the evaluation works
The publisher runs a specified suite and combines its components using documented category weights. Some tasks are scored by checks; others use judges or pairwise assessment.
The current methodology identifies its version and component evaluations. Comparing two index values requires the same recipe as well as comparable model settings.
Configuration matters
Keep reasoning effort, model revision, tool-enabled harness and token limits visible. The Intelligence Index is primarily text and English; separate multimodal measures should not be folded into it.
Primary methods: Intelligence benchmarking methodology ↗
Results at the publisher
Scores stay with the original source
We have not documented the permissions needed to reproduce this leaderboard here. You can still explore its methods and visit the publisher for results.
Versions & freshness
Verified 15 October 2026v4.3.2 is the methodology displayed when reviewed. Earlier index versions used different components; compare values only within their stated methodology version.
Official site and methodology accessible. Models and methodology are updated as the publisher releases evaluations; no fixed schedule is promised here.
Our review date is not an evaluation date. Individual run dates are not consistently available; they remain unknown here. No automatic result collection is enabled.
Publication & reuse
The website terms do not provide the commercial reproduction, storage or derivative-data grant needed for this directory. No separate agreement is documented. Original commentary and links are distinct from copying the publisher's content.
No software licence is being used as permission for leaderboard data.
View the scope of the reuse decision
- Result fields
- No result fields cleared
- Public result display
- not enabled
- Storage
- not enabled
- Historical copies
- not enabled
- Updates
- not enabled
- Commercial reuse
- not enabled
- Derived ratings
- No scoped grant documented; not enabled.
- Attribution
- Credit the named publisher and link the primary sources. Attribution alone does not grant reproduction rights.
This decision does not clear mixed-source scientific recommendations. Publisher content, software, datasets and model licences are separate. No partnership is implied.
Primary sources & attribution
This guide is original scor.es commentary about Artificial Analysis's work. The links below support the identity, methods, versions and reuse review.
- Official website ↗Checked 15 October 2026
- Intelligence benchmarking methodology ↗Checked 15 October 2026
- Website terms of use ↗Checked 15 October 2026
Editorial edition 2026-10-15.1 · reuse review directory-rights-1. Re-check due 14 November 2026.