ctxd - 2026 Pricing, Features, Reviews & Alternatives

ctxd is a cloud-based context infrastructure platform built for developers constructing AI applications and agents. It connects to Google Drive (Docs, Sheets, Slides, PDFs, and images via OCR), Slack (messages, threads, and file attachments), GitHub (repositories, issues, pull requests, and code files), and Google Calendar (events and scheduling data). Every connected file is automatically parsed, chunked, and indexed, then made queryable through four interfaces sharing a single index: a Python SDK, CLI, REST API, and a remote MCP endpoint. No embedding or reranking model is required; search runs directly against indexed text and file paths using a query DSL, eliminating the need for a vector database or chunk-size tuning.

ctxd also ships as an open-source local mode called ctxfs, installable via PyPI, which tracks selected folders on the user's own storage and compute, stores text with source paths, and publishes changes automatically without re-embedding or an API key. The hosted service is available across four tiers: Free (1 user), Pro ($15/user/month), Team ($25/user/month, with SSO and admin controls), and Enterprise (custom pricing, unlimited documents, SAML, SLA, and on-premises deployment). The MCP server exposes three tools to any MCP-compatible AI agent: search, fetch_document, and get_profile, which returns connected integrations and a hierarchical file tree. Agents receive source text and file paths alongside each result, supporting iterative search refinement and evidence citation. The model and agent loop remain external and fully user-controlled.