Artificial Analysis Intelligence Index v5
Artificial Analysis's nine-eval model benchmark composite, now 60% private test data, plus a new Progress Index v1.
Overview
Artificial Analysis published Intelligence Index v5 on October 27, 2026. It combines nine evaluations in four equally weighted categories: Knowledge Work (AA-Briefcase, AutomationBench-AA, GDP.pdf), Coding (FrontierCode 2.0, Terminal-Bench 4.0), Science (Terminal-Bench-Science, CritPt v2 hard subset) and General (AA-Omniscience v1.1, HLE-Diamond). Relative to v4.3 it adds FrontierCode 2.0 and Terminal-Bench-Science, removes GDPval-AA v2.1, SciCode and AA-LCR, and moves 60% of the index to private or held-out test data. At launch Claude Opus 5.5 (max) scores 58, GPT-6 Astra 57 and GPT-6.1 Sol 55; Google later reported 59 for Gemini 4 Carbon. A separate Progress Index v1 tracks capability over time on one scale.
This is a sourced directory entry. Features and capabilities are reported in the linked sources; this entry does not document independent Aymorphic testing.
Key features
- Nine evals in four categories, 25% weight each
- New FrontierCode 2.0 and Terminal-Bench-Science; removed GDPval-AA v2.1, SciCode, AA-LCR v1.1
- 60% of the index uses private test sets or held-out answers
- Launch scores: Claude Opus 5.5 58, GPT-6 Astra 57, GPT-6.1 Sol 55, Kimi K3.1 Pro 46
- Companion Progress Index v1 for capability over time
Use cases
- Comparing frontier and open-weight models on one composite score
- Checking cost per task and output-token usage across top models
Pricing
Public index and articles are free to read; paid tiers unknown.
Pricing reflects the source record and may change. Confirm current terms with the provider.
Dates and sources
- Released
- 27 Oct 2026
- Record updated
- 29 Oct 2026
- Record last checked
- 29 Oct 2026