Amp puts AI agents to work beyond code with OpenAI models

Amp is developing a frontier AI agent designed to help software developers move from initial ideas to working products. According to CEO Quinn Slack, OpenAI models account for the majority of tokens used in Amp because of their transparency and clear safeguards. To address the constraint of token costs, Amp launched a Sign in with ChatGPT integration. This feature provides users with access to more tokens and models such as GPT-6 Astra, which Slack notes gives users the freedom to let agents handle larger tasks without counting every token. Amp's internal 20-person team utilizes these agents for various tasks beyond coding, including marketing, writing documentation, and designing pages. In addition to core coding assistance, Amp recently announced Puck, a conversational AI assistant built into the platform that helps users launch and manage other agents. Powered by GPT-Live-1, Calling Puck allows users to interact with the assistant in real time, even from a watch while away from their computers. Furthermore, Amp has been testing an Ultrafast mode with GPT-6 Astra to increase response speeds, which Slack believes will change how users juggle multiple tasks and allow agents more time to verify their own work. Looking ahead, Amp plans to incorporate more pre- and post-coding work into its agents, enabling builders to carry their ideas further with less manual intervention.