Plaid speed is now available for modes using GPT-6 Astra, enabling inference requests up to 6× faster at 6× cost per token via OpenAI's ultrafast tier. Enable it by selecting a compatible mode in the new thread dialog and toggling the speed switch to Plaid. Plaid works only with Amp-provided inference; subagents and non-Plaid models fall back to fast or standard speed, and you can toggle it on or off for existing threads.