Breaking

Claude Sonnet 5.5 Reaches No. 2 in AI Benchmarks, Outpacing OpenAI's GPT-6 Astra

September 28, 2026 at 21:16 EDT

Foundation Models
AI Agents

Anthropic's newly released Claude Sonnet 5.5 has claimed the second position on the Artificial Analysis Intelligence Index, scoring 56 points and surpassing OpenAI's GPT-6 Astra by three points, while only Anthropic's own Claude Opus 5.5 ranks higher at 58.

AI Benchmarks • September 2026

Sonnet 5.5 Surpasses GPT-6 Astra

Anthropic's latest model scores 56 on the Artificial Analysis Intelligence Index, edging OpenAI's GPT-6 Astra with only Opus 5.5 ranking higher.

Intelligence Index Rankings

Opus 5.5

Sonnet 5.5

GPT-6 Astra

Token Cost

$2 / $10

Input / Output per 1M

Output Tokens (Max)

~193K

Per task at maximum effort

Key Trade-off

Sonnet 5.5 generates 60% more output tokens than Opus 5.5 and 7× more than GPT-6 Astra, delivering superior reasoning at higher token cost for intensive use cases.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.