Vibeleaderboard
← All Intel
Intel / post

Exa plugin brings web-scale retrieval to Codex

Source
x.com
Date

Give ChatGPT Work and Codex access to 100B+ websites, papers, docs, and more with the new @ExaAILabs plugin https://t.co/JAp8UiVxP1

Context

The Exa plugin lets ChatGPT Work and Codex, OpenAI's coding , call Exa's web search engine in the middle of a task. OpenAI Developers announced the plugin with a reach of more than 100 billion websites, papers and docs. That is the advertised size of what Exa searches, not how much any single request reads.

According to Exa's setup guide, the plugin bundles two pieces. One is Exa's hosted server; MCP (Model Context Protocol) is a standard way for an AI app to call an outside service's tools. The other is a set of Exa skills, written instructions that tell the agent how to carry out research workflows. The tools cover three jobs. Search runs a plain-language query and returns the content of the best pages rather than a list of links. Read fetches a specific page the user points to, such as documentation, a changelog or an issue. Research runs several searches and returns an answer with citations.

For Codex, this means an agent working against a library can pull in its current documentation or a specific issue thread while it works. Setup happens inside OpenAI's plugin directory: the user installs Exa there and signs in to an Exa account when prompted. Skills load only in sessions started after installation. Exa's guide does not state pricing, so any per-search cost has to be checked with Exa.

Terms in this piece · Glossary
  • AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
  • MCP — The Model Context Protocol — an open standard that lets any AI assistant plug into any tool or data source without custom integration code.
More from OpenAI Developers
Recommended reads
Comments

Checking sign-in…

Loading comments…