Anthropic says OpenClaw-style Claude CLI usage is allowed again
docs.openclaw.ai
April 21, 2026
4 min read
🔥🔥🔥🔥🔥
65/100
Summary
Anthropic's OpenClaw supports both Anthropic API keys and Claude CLI reuse, allowing for standard API access and usage-based billing. Existing legacy Anthropic token profiles are honored at runtime if already configured.
Key Takeaways
- Anthropic provides access to the Claude model family via an API and Claude CLI in OpenClaw, supporting both API keys and legacy token profiles.
- OpenClaw supports prompt caching for the Anthropic API, with default cache settings of 5 minutes for API key authentication.
- Anthropic's 1M context window feature is in beta and must be explicitly enabled per model in OpenClaw.
- The Claude CLI backend is supported in OpenClaw, allowing for sanctioned reuse of the CLI unless new policies are published by Anthropic.
Community Sentiment
Positives
- Codex is perceived as significantly better than Claude Code, offering superior performance and value for users, which may drive more developers to choose it over Anthropic's offerings.
- The restoration of OpenClaw-style usage suggests a potential return to normalcy for users, which could alleviate some concerns about service disruptions and policy changes.
Concerns
- Anthropic's inconsistent communication and policy changes have led to user frustration, causing many to switch to competitors like Codex for more reliable service.
- The recent price increases and reduced limits from Anthropic are damaging their reputation and goodwill among users, prompting concerns about the sustainability of their offerings.
- Users express skepticism about Anthropic's reliability, fearing that future policy changes could disrupt their access to services, which undermines trust in the platform.