Claude Sonnet 5.5
Anthropic releases Claude Sonnet 5.5, 30% faster and cheaper than Sonnet 5 with better benchmarks
“runs 30%+ faster, and costs up to 30% less for most work”
Anthropic launched Claude Sonnet 5.5, which matches or beats Sonnet 5 on benchmarks while being faster and cheaper to run at the same price point. Notably, it is now the model powering claude.ai's free tier, giving Anthropic a stronger free offering than OpenAI's ChatGPT free tier (currently on Luna 5.6). A known 'max' thinking mode bug causes runaway token usage, mirroring a similar issue in Opus 5.5.