Anthropic added max_tokens support to the advisor tool to cap advisor output per call.

Builders can limit advisor responses when full-length output adds latency and output token cost.

For builds using the advisor tool, the cap is set on tools[].max_tokens in the tool definition.

• Jun 12, 2026Anthropic publicly documented stop_details on refusal responses, with category and explanation fields for routing.
• Jun 12, 2026Claude Opus 4.8 defaults the effort parameter to high across all surfaces.
• Jun 12, 2026Claude Opus 4.8 lowers prompt caching’s minimum cacheable prompt length to 1,024 tokens versus Opus 4.7.

This is the world we build in. It moves this fast. We move with it.