Faster Deep & Rush

Amp announced that its 'deep' and 'rush' modes now deliver first tokens 87% faster and overall responses 32% faster, achieved primarily through websocket communication with OpenAI and a recent platform rebuild. The improvements yield up to 40% end-to-end speedups on long-horizon tasks.

The first token now arrives 87% faster and entire responses are 32% faster, p50, in Amp's deep and rush modes. How? Mostly using websockets for communication with OpenAI, partly because we rebuilt Amp https://ampcode.com/news/neo to be much faster last month. These gains matter most on long-horizon tasks, where we're seeing up to a 40% end-to-end speedup from user prompt submission to completion. You can see the difference: