Anthropic Unveils Opus 4.6 Fast Mode: 2.5x Lower Latency

Anthropic has introduced a fast mode for Claude Opus 4.6 that promises up to 2.5x faster responses as an early experiment. Available via Claude Code (/fast) and the API, it targets urgent tasks but comes with higher operational costs and no pricing yet.

Anthropic Unveils Opus 4.6 Fast Mode: 2.5x Lower Latency

TL;DR

  • Opus 4.6 fast mode — 2.5x faster: Performance-optimized execution mode targeting substantially lower latency (https://anthropic.com/news/opus-fast-mode)
  • Access via Claude Code (invoke with /fast for accounts with extra usage) and an API early experiment
  • Research-preview availability on partner platforms: cursor_ai, emergentlabs, FactoryAI, figma, github Copilot, Lovable, v0, windsurf
  • Higher operational cost; positioned for urgent or high-stakes tasks where faster responses are a priority

Anthropic has introduced a fast mode for Claude Opus 4.6, described as a 2.5x faster variant and available as an early experiment; details are published at the company blog: Opus fast mode.

What’s new

This release centers on a performance-optimized execution mode of Opus 4.6 that aims to deliver substantially lower latency. The build is presented as 2.5x faster than the base Opus 4.6 configuration while preserving the model’s intelligence level. The fast mode is positioned for tasks needing speed and is noted to have a higher operational cost compared with standard modes.

Access and interfaces

Fast mode is being exposed initially through two channels:

  • Claude Code, where fast mode is available now for users who have extra usage enabled (invoked with /fast).
  • The API, provided as an early experiment for developers and integrators.

Availability across partners

Anthropic also listed a set of partners and platforms where the fast mode is appearing in research previews. Those include Cursor, FactoryAI, Figma, GitHub Copilot and Windsurf

Cost and intended use

The build is explicitly characterized as more expensive to run, and Anthropic frames the mode for scenarios described as urgent or high-stakes where faster response times are a priority. No pricing or broader rollout schedule was included in the announcement.

For the original announcement, see the source: https://x.com/claudeai/status/2020207322124132504

Continue the conversation on Slack

Did this article spark your interest? Join our community of experts and enthusiasts to dive deeper, ask questions, and share your ideas.

Join our community