// AI Platform Drift

Claude Sonnet 5.5 released with five breaking API changes

Claude Platform release notes · published 2026-09-29 04:40 UTC
ACTION NEEDED MIGRATE APIMODEL
WHAT CHANGEDClaude Sonnet 5.5 (claude-sonnet-5-5) launched September 28, 2026 on Claude API, AWS, Google Cloud, and Microsoft Foundry. Five breaking changes from Claude Sonnet 5: thinking parameter syntax changed from "disabled" to {"type": "between_tools"}, forced tool_choice types (any and tool) now return 400 errors, thinking blocks are conversation-scoped, computer_20251124 tool rejected, advisor tool rejects Claude Opus 4.8/4.7 and Sonnet 5 as advisors. Thinking blocks from Sonnet 5.5 work only in the originating account or linked accounts; other accounts have blocks silently dropped.
WHY IT MATTERSExisting code calling Claude Sonnet 5 will break if pointed at Sonnet 5.5 without changes. The thinking parameter change breaks any code disabling up-front thinking. The tool_choice change breaks any code forcing tool use. The cross-account thinking block restriction breaks any workflow that passes conversation history between unlinked accounts, though the request succeeds silently with degraded context.
WHAT TO DOBefore upgrading any integration to claude-sonnet-5-5: audit code for thinking: "disabled" and replace with thinking: {"type": "between_tools"}, remove tool_choice values of "any" or "tool", verify computer use calls reference a newer tool version than computer_20251124, check advisor tool calls do not specify rejected models, and confirm all accounts sharing conversation history are linked if using Sonnet 5.5 thinking blocks. Review the migration guide at the release notes URL for request examples.
Model-reported confidence 95% (not measured accuracy) Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
+## Claude Platform release notes — September 28, 2026
+https://platform.claude.com/docs/en/release-notes/overview#september-28-2026
+Mon, 28 Sep 2026 00:00:00 GMT
+
+We've launched Claude Sonnet 5.5 (claude-sonnet-5-5). It's available on the Claude API, Claude Platform on AWS, Claude on Google Cloud, and Claude in Microsoft Foundry. For its context window, output limits, and prices, see the Claude Sonnet 5.5 model page.
+Code written for Claude Sonnet 5 can break on Claude Sonnet 5.5 in five ways. To turn off up-front thinking, send thinking: {"type": "between_tools"} instead of "disabled", at high effort or below. Forced tool use (tool_choice types any and tool) returns a 400 error. Thinking blocks are tied to the model and the conversation. The earlier computer_20251124 computer use tool isn't accepted. The advisor tool rejects Claude Opus 4.8, Claude Opus 4.7, and Claude Sonnet 5 as advisors. See What's new in Claude Sonnet 5.5 for each change and the migration guide for before-and-after requests. For model-specific prompting patterns, see Prompting Claude Sonnet 5.5.
+Thinking blocks that Claude Sonnet 5.5 produces work only in the account that produced them, or in an account linked to it. When another account sends one of these blocks, the API drops the block before the model sees it, and the request succeeds. Blocks from earlier models aren't affected. See Preserved thinking.
HOW IT WAS JUDGED
Triage signals below are model-reported assertions, preserved verbatim; they are not independent arithmetic verification.
triage verdictMATERIAL at 0.95, needed 0.45
triage reasoningNew model with multiple breaking changes and data handling restrictions requires engineering attention.
signalsnew model released (Claude Sonnet 5.5) · breaking API changes documented (thinking parameter syntax, tool_choice behavior, tool compatibility) · model availability across multiple platforms · migration guide required · cross-account thinking block restrictions (data handling change)
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)