Source record

Published by Amp News (ampcode.com) (T1 source). The original is at https://ampcode.com/news/plaid-mode.

Pipeline notes

The summary and note below are generated by the signal pipeline — they are Beyond Desk’s reading, not quotations from the source.

Amp introduced a Plaid speed option for modes using GPT-6 Astra, offering inference up to 6x faster at 6x the per-token cost. Users enable it via the mode picker by moving the speed switch past Fast. Subagents and non-Plaid models fall back to fast or standard speed when Plaid is active.

Concrete speed-vs-cost toggle in Amp, with the exact steps and subagent fall-back behavior spelled out. Useful for timing-sensitive coding sessions if the cost math works.

Cited by

No citations on record.