OpenAI spotlights Amp as example of agents using GPT-6 Astra, GPT-Live and Sign in with ChatGPT

OpenAI on Wednesday highlighted startup Amp as a concrete example of how third-party apps are using its newer tools to push AI agents beyond software coding and into broader workplace tasks. In a post published Oct. 14 titled “Amp puts AI agents to work beyond code with OpenAI models,” OpenAI profiled the company as an adopter of its latest models, voice features and sign-in system.

The products at the center of the profile are GPT-6 Astra, GPT-Live-1, OpenAI’s low-latency Ultrafast service tier and Sign in with ChatGPT. OpenAI said Amp is using GPT-6 Astra not just for coding, but also for writing, design and other non-code tasks, while the roughly 20-person company also uses agents internally for marketing and product work. Amp CEO Quinn Slack said in OpenAI’s post, “We have a very unique design style and writing style. We’re very picky about this, but we’ve been really happy with how GPT-6 Astra does.”

The most important operational detail is the Sign in with ChatGPT integration that OpenAI said Amp has launched. Sign in with ChatGPT is OpenAI’s sign-in and authorization product for third-party apps; according to OpenAI’s documentation, it lets users sign in with their ChatGPT identity and, when eligible, use their ChatGPT plan for in-app requests instead of supplying a separate API key. That matters for an agent product like Amp because it shifts some of the billing and access friction away from the user. In OpenAI’s telling, Amp attributes larger agent workloads to that integration, and OpenAI’s post says Amp reports that some users consume “10 billion tokens a week.” Slack said, “Sign in with ChatGPT means that people can have that feeling where they’re not counting every single token.”

Amp pitches itself as a developer-focused agent product built around persistent agents that keep running after a user steps away. Its website describes those agents as “orbs,” alongside a built-in assistant called Puck, portals and mobile apps for monitoring and controlling agents. OpenAI said Amp announced Puck as a conversational assistant that helps launch and manage agents, and that Amp also supports “Calling Puck,” powered by GPT-Live-1, for real-time voice interactions, including from a watch. OpenAI also said Amp has been testing Ultrafast mode with GPT-6 Astra. Slack, again in OpenAI’s post, said, “People have not seen a model that’s this good, that’s this fast.”

The underlying OpenAI products are not new today. GPT-Live was introduced July 8, 2026, Ultrafast was previewed Aug. 13, 2026, and OpenAI’s GPT-6 Astra page was updated in September. What is new in Wednesday’s post is the implementation example: OpenAI is pointing to Amp as an early, real-world case of identity, low-latency response and voice interaction being combined into heavier-duty agent workflows, rather than announcing another model release.

There is also a brief safety backdrop to Astra. OpenAI has published materials discussing Astra’s cyber capabilities and safeguards, and The Associated Press recently reported that OpenAI delayed a follow-up release, GPT-6.1 Astra, over security concerns. In the OpenAI post, Slack said the company values that approach: “OpenAI is very transparent, and when there are safeguards applied, users know and they understand. That makes it really easy to use OpenAI models. You can do all of your work in Amp with OpenAI models and be really happy.”