The Messages API response now includes usage.output_tokens_details.thinking_tokens, reporting how many of the billed output tokens were extended thinking. When streaming, the breakdown appears only on the final message_delta event. No beta header is required.
Fetched June 6, 2026
The advisor tool now supports a parameter to cap the advisor model's output per call, reducing latency and output token cost for workloads…
Anthropic · Claude PlatformClaude Opus 4.8 is now available as the most capable generally available model. Key features and updates: - 1M token context window by defa…
Anthropic · Claude PlatformThe Copilot usage metrics API now reports an aicreditsused field per user per day in the user-level reports. The field is available in both…
GitHub · GitHub ChangelogChatGPT Enterprise admins can now view credit usage trends across users, products, and models in the Global Admin Console, and set default,…
OpenAI · Product Releases
