Skip to content

An editorial team and a community building with AI

vibecoding.incJoin Discord ↗
← All articles

News · Anthropic · 2 min read

Event date: · Published:

Anthropic released Claude Sonnet 5.5 at Sonnet 5's price, with breaking API changes

On September 28, 2026, Anthropic released Claude Sonnet 5.5 at the same $2 and $10 per million token list price as Sonnet 5. The release also changed API behavior in ways that could break code written for Sonnet 5.

Archive entry: this covers an event from September 28, 2026, and we published it on November 8, 2026. The details come from sources dated at the time and may have changed since.

On September 28, 2026, Anthropic released Claude Sonnet 5.5 (claude-sonnet-5-5), the second model in the Claude 5.5 family. Anthropic described it as a faster, lower-cost complement to Claude Opus 5.5, and it kept the list price of Sonnet 5.

What changed

  • Price: $2 per million input tokens and $10 per million output tokens, the same list price as Sonnet 5, with cache reads at $0.20 per million tokens at launch.
  • Speed and cost: in its testing, Anthropic said output generation was more than 30% faster and cost per task up to 30% lower.
  • Benchmarks: Terminal-Bench 4.0 reached 70.6% against 10.3% for Sonnet 5, CursorBench 4.0 reached 55.5% against 34.1%, and OSWorld 2.1 reached 80.1% against 57.0%.
  • Effort: on several benchmarks, Sonnet 5.5 at Low or Medium effort beat Sonnet 5's best score for about a tenth of the cost per task.
  • Availability: the Claude API, Claude in Amazon Bedrock, Claude Platform on AWS, Claude Models on Google Cloud (Google managed) and Claude in Microsoft Foundry.

API changes that could break existing code

  • To turn off up-front thinking, requests set the thinking type to between_tools instead of disabled, at high effort or below.
  • Forced tool use (tool_choice types any and tool) returned a 400 error.
  • The earlier computer_20251124 tool was not accepted on the Claude API and Google Cloud.
  • The advisor tool rejected Claude Opus 4.8, Claude Opus 4.7 and Claude Sonnet 5 as advisors.
  • Thinking blocks from Sonnet 5.5 worked only in the account that produced them or a linked account; the API dropped blocks from other accounts and the request still succeeded.

What it meant for builders

Sonnet was the default workhorse in many coding tools, so a large benchmark jump at an unchanged price mattered. Code that turned thinking off or forced a tool call needed migration work. Claude Code v2.1.284 made Sonnet 5.5 its default Sonnet model on the Anthropic API, and Anthropic said Claude Haiku 5.5 would join the family in the weeks after launch.

Sources

Written with AI assistance, strictly from the sources listed above. Before it went live, automated checks verified its dates, sources and every language edition.

Anthropic released Claude Sonnet 5.5 at Sonnet 5's price, with breaking API changes | vibecoding.inc