Releasesby GlyphClaude API input_transformations adds thinking_mismatch_allowed entries in beta
Under the thinking-binding-controls-2026-08-01 beta header, input_transformations can report thinking_mismatch_allowed for thinking blocks that failed an unenforced prefix check. Logging these entries helps find conversation history edits in production traffic.
Also shipped Sep 14, 2026Anthropic in September 2026 →
Metadata for BigQuery Graph is automatically ingested and searchable in Knowledge Catalog
Metadata for BigQuery Graph is automatically ingested and searchable in Knowledge Catalog . This feature is available in preview .
BigQuery supports WHERE clauses inside aggregate function calls in preview
Aggregate function calls can now include a WHERE clause that filters their input with a boolean expression, simplifying conditional aggregation in SQL.
Multimodal AI_SUMMARIZE for automatic theme summarization (Public Preview)
Snowflake Cortex AI_SUMMARIZE is now available in public preview, enabling customers to summarize text and multimodal content, including images and documents, directly in Snowflake.
Zero-copy support for all table formats in interactive warehouses (General availability)
Support for all table formats in interactive warehouses is now generally available and is no longer in Preview .
Genie visualizations in Slack replies (Public Preview)
When a Genie response includes a chart, the visualization now appears as an image in the Slack reply. If you installed the Databricks Genie app for Slack previously, reinstall it to enable this. See Visualizations .
Metadata for BigQuery Graph is now automatically ingested and searchable in Knowledge Catalog
Metadata for BigQuery Graph is now automatically ingested and searchable in Knowledge Catalog. This feature is available in preview . For more information, see Knowledge Catalog overview .
More from Claude PlatformEvery Claude Platform release →
Messages API adds on-demand conversation compaction in beta
With the compact-2026-09-04 beta header, the top-level compaction parameter returns a signed compaction block summarizing prior messages, which you send in their place on later requests. You choose when to compact, it can run in the background, and recent turns can be kept verbatim.
Compliance API adds Claude in Chrome session transcripts in beta
Compliance API local session endpoints now return transcripts of Claude in Chrome sessions (product_surface claude_in_chrome) for Claude Enterprise organizations, using the existing Compliance Access Key and read:compliance_user_data scope.
For cache diagnostics, a response to a request that sends the cache-diagnosis-2026-04-07 beta header now
For cache diagnostics, a response to a request that sends the cache-diagnosis-2026-04-07 beta header now always includes the diagnostics field. The field is null when the request did not include the diagnostics object. Previously the field was omitted in that case.
Claude Managed Agents permission policies add an auto mode
A new auto permission policy lets the server evaluate each agent or MCP tool call and run it, deny it, or pause for approval. agent.tool_use and agent.mcp_tool_use events now report how each call was evaluated.
ant CLI 1.32.0 adds sessions connect for live Managed Agents sessions
ant beta:sessions connect attaches a terminal to a Claude Managed Agents session so you can follow it live, send messages and approve or deny pending tool calls. --web serves the Console session viewer locally.
Smart reports (beta)
Smart reports provide analytics on team usage, costs, friction points, and potential shared skills. This feature is available in beta for Claude Enterprise plans.
Claude Opus 5.5 launches with a 1M token context window and 128k output
Anthropic released Claude Opus 5.5 for long-running agentic coding and knowledge work, with a default 1M token context, 128k max output and always-on adaptive thinking. It costs $4 / $20 per MTok, below Claude Opus 5, and is on the Claude API, Bedrock, Google Cloud and Microsoft Foundry.
Claude Opus 5.5 breaking changes: thinking can't be disabled, forced tool_choice removed
On Claude Opus 5.5, disabling or manually enabling thinking returns a 400; control depth with the effort parameter instead. tool_choice any and tool also return 400, so use auto with strict tool use, and computer use requires the computer_toolset_20260801 toolset.