Claude Code returns blank thinking blocks, but reasoning still costs you
- ID
- 14116
- Status
- summarized
- Published
- 14 Aug 2026, 3:30 PM
- Fetched
- 14 Aug 2026, 4:16 PM
- Provider
- The Register
- Category
- technology
- Original URL
- https://www.theregister.com/ai-and-ml/2026/08/14/claude-code-returns-blank-thinking-blocks-but-reasoning-still-costs-you/5287557
- Source URL
- https://www.theregister.com/headlines.atom
Summary
- Score
- 7.0
- Created
- 14 Aug 2026, 4:16 PM
- Tags
- Audience
- developersai_agent_usersai_ml_learnersvibe_coders
What happened
Developers report that Anthropic's Claude API returns empty or truncated thinking blocks for Opus 4.8 and Sonnet 5 even when summarized thinking is explicitly requested, yet the underlying reasoning tokens are still billed in full as output tokens. Anthropic's own documentation confirms users are charged for all thinking tokens generated regardless of whether the thinking text is returned, and the issue may stem from display-summary testing rather than a broad outage.
Why it matters
If you enable thinking on Claude Opus 4.8 or Sonnet 5, you may be paying for reasoning tokens you never see — so audit your token usage and consider lowering the thinking budget or disabling thinking entirely until the blank-block behavior is confirmed fixed. This directly affects your API cost per call.
Discussion angle
Compare the real cost-per-task of 'thinking' models versus non-thinking models now that billed-but-undisplayed tokens are a known risk — is the quality uplift worth the opaque billing?