[BUG] Missing Thinking Summary Blocks in the Claude AI Web Interface

Status Closed — duplicate
Reported on v2.1.218
Maintainer reply None cached
Activity 0 comments · opened Jul 23, 2026 · closed Jul 23, 2026

Preflight Checklist

  • [x] I have searched existing issues and this hasn't been reported yet
  • [x] This is a single bug report (please file separate reports for different bugs)
  • [x] I am using the latest version of Claude Code

What's Wrong?

Date/Time Observed: July 23, 2026, at approximately 5:00 AM UTC

Platform/Environment: Claude AI Website

Models Affected: All Models

Description

The thinking summary blocks disappeared from the chat interface starting on July 23, 2026. Previously, these blocks appeared before the generated text to show the reasoning process. Currently, no reasoning or thinking summaries are visible during or after generation, regardless of the model selected.

What Should Happen?

Expected Behavior

The interface displays an expandable thinking block showing the reasoning summary before the final response appears.

Actual Behavior

The chat interface goes into thinking in the background without showing a summary of what the model was thinking. The thinking summary block is completely missing from the screen.

Error Messages/Logs

Steps to Reproduce

Steps to Reproduce

  1. Log into the Claude AI web interface.
  1. Select any model with reasoning capabilities (Sonnet 4.6, Sonnet 5, or Opus).
  1. Send a message to the chat.
  1. Watch the response generation.

<img width="720" height="799" alt="Image" src="https://github.com/user-attachments/assets/ac109149-4ccd-47e6-a975-f8ad75fb6db2" />

Claude Model

Not sure / Multiple models

Is this a regression?

Yes, this worked in a previous version

Last Working Version

_No response_

Claude Code Version

2.1.218

Platform

Other

Operating System

Windows

Terminal/Shell

Windows Terminal

Additional Information

Use Case and Impact

I use the thinking block summaries to read how the model interprets my questions in real time. Seeing the initial reasoning process allows me to catch misunderstandings early. I can then stop or edit my prompt before the text generation finishes. The absence of this feature removes my ability to course-correct, leading to wasted time and causing message limits to be used up faster through multiple rerolls.

View original on GitHub ↗