September 29, 2026
Claude Code Will Handle More Tasks: Sonnet 5.5 Promises +30%
On Sep 28, Anthropic added Sonnet 5.5 to its CLI: in the company's tests, the model completed 30% more tasks than Sonnet 5. In the Anthropic API, Sonnet 5.5 became the default Sonnet and gained a 1 million-token context window. Pricing is $2 per 1 million input tokens, $2 for cache writes, and $10 for output tokens. Cache reads cost $0.20.
cat
@_catwu
Sonnet 5.5 is here! In Claude Code, users complete about 30% more tasks than with Sonnet 5. The model is smarter, so it needs fewer tokens for the same work. Here, it clears leaves from a yard using tool calls. It completes the task 24 seconds faster than Sonnet 5 and uses 6 thousand fewer tokens. That leaves it a little extra time to read :)
Introducing Claude Sonnet 5.5, the second model in the Claude 5.5 family. It’s a clear upgrade over Sonnet 5, runs more than 30% faster, and costs up to 30% less for most work.
· 12.7K views
Sonnet 5 previously used more tokens for the same work. According to Anthropic, Sonnet 5.5 does it faster and with lower usage: in one run with tool calls, the difference was 24 seconds and 6 thousand tokens.
The model is already in the CLI. For a single session, launch Claude Code with `claude --model claude-sonnet-5-5`. In an already open CLI, select the model through `/model`.
Claude Code 2.1.284, released on Sep 28, already includes Sonnet 5.5.
Source
