AI Frontier

Frontier Language Model Intelligence, Over Time

A release-date skyline for the models Artificial Analysis marks as frontier language models.

Ranked by AA Intelligence Index Updated July 31, 2026 Daily refresh Intelligence Index Heights
Preparing district view
Intelligence Index Heights Embedded district view
Explore full city

Latest ranking

July 31, 2026

AA Intelligence Index
#1 Claude Opus 5 (Adaptive Reasoning, Max Effort) 60.7
#2 Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) 59.9
#3 GPT-5.6 Sol (max) 58.9
#4 Kimi K3 (max) 57.1
#5 GPT-5.6 Terra (max) 55
#6 Grok 4.5 (high) 53.8
#7 Claude Sonnet 5 (Adaptive Reasoning, Max Effort) 53.4
#8 GPT-5.6 Luna (max) 51.2
#9 GLM-5.2 (max) 51.1
#10 Muse Spark 1.1 (xhigh) 50.6
#11 Gemini 3.6 Flash (high) 50.1
#12 DeepSeek V4 Flash 0731 (Reasoning, Max Effort) 49.9

Timeline

Key moments

  • February 17, 2026

    Claude Sonnet 4.6 (max) sets the frontier

  • June 1, 2026

    Claude Opus 4.8 (max) sets the frontier

  • July 31, 2026

    Claude Opus 5 (max) sets the frontier

Notes

Methodology

This adapter reads the public Artificial Analysis homepage payload, prioritizes rows explicitly marked as frontier models, ranks them by the current Artificial Analysis Intelligence Index, and plots them by release date so the skyline updates when Artificial Analysis publishes new model evaluations.

Observed 14 Artificial Analysis frontier model rows and 14 scored display rows at 2026-08-05T06:06:35.481Z.

Reference

Sources