Stack Overflow for Agents triples signups after launching ChatGPT plugin
Stack Overflow says registrations for Stack Overflow for Agents tripled overnight after it released a ChatGPT plugin, according to a case study published by OpenAI. The company also credits Codex with helping a small team ship the product in a few weeks and with shortening the time it takes to build enterprise connectors.
- Stack Overflow says a ChatGPT plugin tripled registrations for Stack Overflow for Agents overnight.
- Two engineers using Codex built the product in a few weeks, according to Stack Overflow.
- Connector builds dropped from 3 to 6 weeks on average to 4 days with Codex, and one was built in a single day.
- Stack Overflow's platform uses OpenAI models such as GPT-6 Astra and GPT-6 Sol, with routing by reasoning needs and cost.
- The company hopes to explore the Decisions API to reduce model routing and configuration work.
A plugin that removes setup friction
Once Stack Overflow for Agents was ready for distribution, the team needed a way to get it into people's workflows. It built a ChatGPT plugin so users could discover and set up the product inside ChatGPT, with fewer steps than a traditional Model Context Protocol (MCP) setup.
Alexandra Lato, VP of Product at Stack Overflow, put it this way: "We built a ChatGPT plugin for Stack Overflow for Agents, and our registrations tripled overnight." She added that "Once you have a plugin, it becomes a frictionless setup with ChatGPT." According to Lato, the plugin also accelerated agent creation: "I think we were up to 10X last time I checked."
Codex behind the product and the connectors
Stack Overflow is building a knowledge intelligence platform that unifies scattered sources such as Slack and Google Docs and turns raw context into verified knowledge for people and agents. According to OpenAI, models like GPT-6 Astra and GPT-6 Sol power the platform's AI, with tasks routed to different models based on the reasoning required and cost.
Stack Overflow for Agents, which extends the company's intelligence layer to agents, began with a hypothesis and two engineers using Codex. They gave Codex a development environment and a plan, then worked with it to build the product. Lato says they had a full-fledged product within a few weeks. "So it really made their team feel bigger," she says.
Codex is also used on Stack Internal, the company's enterprise product, which must connect to the systems where organizations keep their knowledge. Lato describes a shorter path to market for that product: three months rather than a year. On connectors, she says: "On average, it takes us 3 to 6 weeks to build one connector. With Codex, we were able to build it within 4 days." More recently, the team built a connector in a single day.
Decisions API and the full knowledge lifecycle
Stack Overflow manages its own model routing and configuration. Lato hopes to explore the Decisions API to reduce some of that work and free up engineering capacity for the product. "If we're able to automate that through OpenAI directly, that saves us a ton of time and money," she says.
The company is also looking at how knowledge flows back out. So far, its focus has been pulling information in. Lato is now working on sharing what people learn, so that knowledge gained in a one-to-one session with a developer or someone outside engineering can help others. "As I'm learning, how do I make sure that the rest of my organization is learning with me?" she said.
What it means
The case study, which reflects Stack Overflow's own account, points to distribution inside ChatGPT as a lever for adoption of agent products, since a plugin needs fewer steps than a typical MCP setup. It also shows a company using Codex to shorten build cycles for both a new product and enterprise integrations. The figures on signups, agent creation and connector timelines come from Stack Overflow, not from an independent measurement.