Anthropic has officially released "Fast Mode" for its Opus 5.5 model, moving it out of research preview. This feature doubles the cost per token, with input tokens priced at $8 per million and output tokens at $40 per million, while promising up to 2.5 times faster output token generation. However, the speed benefits are primarily focused on output tokens per second and do not significantly reduce the time to the first token. A critical consideration is that switching between fast and standard modes invalidates the prompt cache, potentially leading to higher costs for subsequent requests. AI
IMPACT This release offers users faster output generation at a premium price, potentially impacting cost-benefit analyses for real-time AI applications.
RANK_REASON General availability release of a specific model version with a new feature (Fast Mode) from a frontier lab. [lever_c_demoted from frontier_release: ic=2 ai=1.0]
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →