Claude Sonnet 5.5 reportedly stealth testing: 1M context, $2/$10 per 1M tokens

kimmonismus · x · 2026-09-24

Unverified leak: Anthropic is reportedly stealth-testing Claude Sonnet 5.5 (claude-sonnet-5-5). Circulating specs: a 1M-token context window, 128K max output, and pricing of $2/$10 per 1M input/output tokens. Commenters read it as a direct response to GPT-6-Sol with matching pricing and presumably comparable or better performance. Not yet confirmed by Anthropic.

Related event: Rumor: Anthropic Quietly Testing Claude Sonnet 5.5 with 1M Context(2 posts)→

Original post →

More from Models

Models channel →