# Anthropic releases Claude Sonnet 5.5

The model runs over 30% faster and reduces per-task costs by up to 30% through improved token efficiency compared to Sonnet 5.

- Priced at the same per-token rate as Sonnet 5 while requiring fewer tokens to complete tasks
- Beats Sonnet 5's top benchmark scores at Low or Medium effort levels for roughly one-tenth the cost
- First Sonnet model to incorporate cyber safeguards and fallbacks found in higher-tier models
- Available immediately across all Claude platforms, with Claude Haiku 5.5 planned for the coming weeks

## Why it matters

Developers and teams get faster output and lower operational costs for everyday coding, document creation, and iterative tasks.

## Sources

- [Anthropic: Anthropic releases Claude Sonnet 5.5](https://x.com/claudeai/status/2104633115620823187)

---

Summarized by dstilled on 2026-09-28. https://dstilled.ai/story/d3d5be87-5820-4d68-a534-1abcf887699d