Upstash

Context7 gives AI coding agents up-to-date documentation and code examples for software libraries. Instead of relying on what a model learned during training, an agent can look up the current docs for the version of a library a developer is using.

The site says it works with Claude, Codex, Cursor and other agents. Developers install it into their agent and then ask questions about a library, such as how to call a function or set up a feature. Context7 returns matching documentation and examples for that library.

The homepage lists public documentation sources such as Next.js, React, Playwright, PostgreSQL, Supabase, GitLab and Cloudflare, with each source's benchmark score, number of code snippets and how recently it was updated. Teams can also add their own docs, and the site has a Try Live page and an install guide.

The Free plan costs $0 and includes access to public repositories, basic access control and 1,000 API calls per month. Free users who reach the limit get 20 bonus calls a day until the month resets. The Pro plan costs $10 per seat per month, with 2,000 API calls per seat and private repositories, team collaboration and unlimited API calls at $5 per 1,000 calls beyond the allowance. Parsing private repositories is charged separately at $5 per 1M tokens.

Enterprise pricing is custom. The pricing page says small teamspaces start at $30 per user per month and larger ones can go as low as $2.50 per user per month, and it lists SOC-2, SSO and a self-hosted option. Teamspaces of up to 50 members can sign up for Enterprise without talking to sales.

The site states that only the documentation lookup query and the library name are sent to Context7's servers, not the user's code or conversation history. It also says the company runs an injection-detection model on stored documentation and assigns trust scores to repository organizations.

Context7 is an Upstash project. The site links to its GitHub repository and to its X account.