Voices
Opus 4.6 Fast Mode Launches on Cloudflare AI Gateway — 2.5x Faster Output Tokens, Same Model Intelligence
Cloudflare announced Opus 4.6 fast mode on AI Gateway, delivering 2.5x faster output token speeds while maintaining identical model intelligence — not a smaller model, but the same Opus 4.6 with optimized serving. Developers can enable it with 'speed: fast' in Anthropic provider options or 'fastMode: true' in Claude Code. This closes the speed gap that historically pushed latency-sensitive applications toward smaller models, potentially shifting the default for production agentic workflows.
Source
↳ Follow the thread