Skip to content

Anthropic

Fast mode is Claude API only, and limited to three Opus models

Written by an AI model from the linked sources and checked against them by a second model. How we work

All about: Claude prices and limits →
Image: platform.claude.com

Fast mode delivers up to 2.5x more output tokens per second on Claude Opus 5, Opus 4.8, and Opus 4.7, at a premium price. You turn it on with the fast-mode-2026-02-01 header and the speed: “fast” parameter. It’s a research preview for now, with access through an account manager or a waitlist.

For Opus 5 and Opus 4.8, the mode runs only on the Claude API, including Claude Managed Agents. It isn’t on Amazon Bedrock, Google Cloud, or Microsoft Foundry. The model’s intelligence and capabilities are unchanged; only the output configuration is sped up.

Fast mode carries its own rate limit: exceed it and the API returns a 429 with a retry-after header. Fast mode for Opus 4.7 was deprecated on June 25, 2026 and gets removed on July 24. It’s been unavailable on Opus 4.6 since June 29, where requests run at standard speed and standard rates.

No comments yet. Be the first.

Join the conversation

Sign in to leave a comment. It appears after moderation.