Amp puts AI agents to work beyond code with OpenAI models

Amp, a frontier AI agent platform, now supports OpenAI’s newest models including GPT‑6 Astra, Ultrafast, GPT‑Live‑1, and Sign in with ChatGPT. The integration allows developers to expand agent capabilities beyond coding to include planning, design, and post-deployment tasks. Amp’s CEO Quinn Slack emphasizes that OpenAI’s transparency and safeguards make its models a preferred choice for users, accounting for the majority of tokens processed in Amp’s system.

Quinn Slack highlights that some Amp users consume up to 10 billion tokens weekly, noting that token cost concerns can limit experimentation. To address this, Amp introduced Sign in with ChatGPT, which provides additional tokens and access to models like GPT‑6 Astra. This integration reduces the need to monitor token usage closely, encouraging users to assign larger tasks to agents and discover their full potential.

Amp’s 20-person team uses its own agent to automate marketing, improve user onboarding, and enhance search discoverability. The agent relies on high-quality writing and design, tasks where GPT‑6 Astra helps meet Amp’s standards. Quinn Slack also experimented with GPT‑6 Astra at home, using it to generate a 3D model of his house in Blender and later turning it into games with his children.

Amp recently launched Puck, a conversational AI assistant that helps users launch and manage other agents. With Calling Puck powered by GPT‑Live‑1, users can interact with Puck in real time, even from a watch. Amp is also testing Ultrafast mode with GPT‑6 Astra, which Quinn says changes how people use agents by enabling faster responses and reducing the need to multitask between agents.