Artificial Analysis Intelligence Index v4.2 Released

Serdar HocamAuthor & Editor

According to the newly published index report, GPT-6 Astra delivers high token efficiency and demonstrates significant success compared to previous models.

◉ 0 views
Announcing Artificial Analysis Intelligence Index v4.2

The Intelligence Index v4.2 report announced by Artificial Analysis revealed current performance and efficiency rankings of artificial intelligence models.

Performance of the GPT-6 Astra Model

According to the newly published data, GPT-6 Astra exhibits quite high token efficiency compared to almost all other models at the intelligence frontier.

The model provides a significant performance increase and superiority of approximately 85 Elo points compared to the GPT-5.6 Sol model.

AA-Briefcase and Model Rankings

Claude Fable 5.1 and Opus 5 models developed by Anthropic hold the leadership in the AA-Briefcase category.

Following these two models, GPT-6 Astra and Muse Spark 1.3 models take places in the ranking, drawing attention.

Token Usage and GDP.pdf Results

Claude Fable 5.1 and Gemini 3.8 Flash stand out as the options using the most tokens among models scoring 25 and above in the index.

OpenAI is in a leading position in GDP.pdf results with rates of 33.2 percent for GPT-6 Astra and 28.2 percent for GPT-5.6 Sol.