AI BriefingAnthropic - OpenClaw: Usable Again
Key point
The OpenClaw tool for Anthropic's Claude models is supported again, supporting the latest features such as Claude 4.6's adaptive thinking and prompt caching.
Details
With Anthropic's Claude CLI-style usage being allowed again, both an Anthropic API key and the existing Claude CLI can now be used via OpenClaw. Previously configured Anthropic token profiles are also retained as they were.
The Claude 4.6 model applies adaptive thinking by default when no explicit configuration is set, and users can directly control it via the /think:<level> command or model parameters.
The key supported features and configuration details are as follows:
- Fast Mode: The
/fasttoggle can switchservice_tierbetweenautoandstandard_only. However, this only applies when requests go directly toapi.anthropic.comwithout passing through a proxy. - Prompt Caching: An API-only feature that allows cost optimization by setting a cache retention time of either
short(5 minutes) orlong(1 hour). - 1M Context Window: A beta feature that can be enabled via
params.context1m: truein the model settings to use a large-scale context.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
Meta Admits AI Assistant 'Muse' Was Heavily Inspired by OpenClaw