Sign in
← AI PulseSep 22, 2026

Wire · news · Single-source brief

Claude Opus 5.5: Updates for Agentic Workflows

Anthropic has released Claude Opus 5.5, a model designed for long-running agentic coding and knowledge work, introducing several breaking changes for existing applications.

By Illumora Editorial

Source · Sep 22, 2026, 4:30 PM · On Illumora · Sep 22, 2026, 4:32 PM

Media from the primary source — shown here so you can stay on Illumora.

Rewritten from one allowlisted primary — not independent enterprise reporting. Lanes →

Brief drafted by Illumora’s editorial model from the linked primary source. Ops desk reviews flagged pieces. How we write →

Read the source →Anthropic Docs (prompting) — Claude Opus 5.5
Save

Anthropic has introduced Claude Opus 5.5, a new model intended for agentic coding and knowledge work that involves extended operations. This release includes updates to its behavior and pricing structure.

Key Points

  • Claude Opus 5.5 is designed for long-running agentic coding and knowledge work.
  • Pricing for Claude Opus 5.5 is $4 USD per million input tokens and $20 USD per million output tokens.
  • Four breaking changes affect code previously running on Claude Opus 5: thinking cannot be disabled, forced tool use now returns an error, thinking blocks are tied to the model and conversation, and the computer_20251124 computer use tool is no longer accepted on the Claude API and Google Cloud.
  • The first three breaking changes also apply to Claude Fable 5.1.
  • Adaptive thinking is always active in Claude Opus 5.5 and cannot be deactivated; its depth can be controlled with the effort parameter.
  • On the Message Batches API, Claude Opus 5.5 supports up to 300k output tokens when using the output-300k-2026-03-24 beta header.
  • The minimum cacheable prompt length is 512 tokens.

Context

According to Anthropic's documentation, Claude Opus 5.5 offers specific behavioral differences and prompting patterns. The model provides control over thinking depth, latency, and cost, allowing users to select a level appropriate for their workload. A lower-latency version of Claude Opus 5.5 is available as a research preview on the Claude API with separate pricing.

Why It Matters

Builders using previous Claude Opus versions for agentic workflows must review and adapt their code due to the breaking changes in Claude Opus 5.5. The updated pricing and output token limits also affect cost management and application design for long-running tasks.

What To Do

  • Review the Claude Opus 5.5 documentation for detailed information on the four breaking changes.
  • Adjust existing code to account for the new behavior of thinking blocks and tool use, especially if migrating from Claude Opus 5 or Claude Fable 5.1.
  • Compare the new pricing structure for Claude Opus 5.5 against current usage to assess potential cost impacts.
  • Test applications with the output-300k-2026-03-24 beta header on the Message Batches API if extended output tokens are required.
  • Note the 512-token minimum for prompt caching to optimize prompt design.