Anthropic drops the waitlist: Claude Opus 5.5 fast mode is now available to everyone

#What happened

On October 15, Anthropic's release notes took Claude Opus 5.5 fast mode out of research preview on the Claude API. Organizations no longer need to request access through an account manager or a waitlist, and no beta header is required (requests that still include it keep working). Every usage tier now has a default fast mode rate limit for Opus 5.5: Start, Build, and Scale tiers get default limits separate from the standard-speed limit, while Evaluation-tier organizations get it once they move up. Anthropic says it is the same model running on a faster inference configuration, with no change to intelligence or capabilities. It multiplies output tokens per second by up to 2.5, a gain that shows in streaming interfaces rather than time to first token. It costs twice the standard rate: $8 per million input tokens and $40 per million output tokens, versus $4 and $20 for standard Opus 5.5. It cannot be used with the Batch API. It is also available on Claude in Microsoft Foundry, but only for deployments hosted on Azure — not for Anthropic-hosted deployments, AWS, or other partner clouds. Fast mode on Opus 5 and Opus 4.8 remains in research preview ($10/$50) and still needs an access request.

#Why it matters

Fast mode is Anthropic's answer to the agent-workload speed problem: same frontier reasoning, streaming up to 2.5x faster for agentic coding and long tasks where output throughput is the bottleneck. At exactly double the per-token price, it is a pure speed-for-money trade that only pays off when latency matters. For developers already defaulting to GPT-6.1 Sol or Opus 5.5 in agentic tools, this removes the last access friction on Anthropic's fastest option.

Verified 2026-10-16.