SuperAPI AI Directory

Vexa

Vexa is an open-source meeting bot API for Google Meet, Microsoft Teams, and Zoom. It provides real-time transcription, bot control, and meeting data interfaces, supports REST API, WebSocket, and MCP, and can be self-hosted or used via its cloud service. Bot base pricing starts at $0.30 per hour, plus $0.50 per hour for real-time transcription. Users can connect meeting data to MCP-compatible clients such as Claude Code and Codex, letting agents read meeting transcripts in real time; they can also embed a meeting bot into their own product with a single API call. The platform is licensed under Apache 2.0, supports deployment with Docker Compose, Kubernetes, or OpenShift, stores data in PostgreSQL, and provides a fully featured web dashboard for managing bots, viewing transcripts, and monitoring meetings.

Open website

About this tool

Purpose: Vexa turns Google Meet, Microsoft Teams, and Zoom meetings into real-time data that can be used by programs or AI agents. It can be used as a standalone tool to record and transcribe meetings, as an underlying engine embedded in third-party products, or to connect meeting data to existing AI agents.

Key capabilities: Supports real-time transcription, returning text with speaker names via WebSocket while the meeting is in progress, with per-meeting language settings; provides three access methods: REST API, WebSocket, and MCP; includes bot control, audio capture and recording, and event notifications for meeting start/end/failure; all data is stored in PostgreSQL, allowing history queries, conversation search, and export via API. It also provides an MCP toolkit that can connect to MCP-compatible clients such as Claude Code, Codex, and OpenClaw, letting agents read real-time meeting data directly.

Usage: A bot can be sent to join a meeting with a single API call, transcripts can be streamed via WebSocket, and recordings are retained. The individual-oriented solution supports pasting prompts into Claude Code, Codex, or any MCP client, letting agents join meetings and read transcripts. For self-hosting, deploy on your own servers with Docker Compose, Kubernetes, or OpenShift, with transcription units and AI models also running locally. An official full web dashboard is provided to manage bots, view transcripts, and monitor meetings, and can be used as-is or forked for customization.

Use cases: Suitable for teams that need to connect meeting content to AI agents in real time; suitable for developers embedding meeting bots in their own products; suitable for fund allocation, private credit, or wealth management scenarios in finance; and suitable for enterprises with data sovereignty requirements that want meeting recordings and transcripts to remain on their own servers and use their own models.

Limitations and pricing: The base rate covers audio capture, webhooks, and 12 months of audio storage; real-time transcription costs an additional $0.20 per hour, for a total of $0.50 per hour. New accounts receive $2 in bot credit (no credit card required), equivalent to about 6 hours of bot time at $0.30 per hour; the minimum top-up is $10. The personal edition is $12 per seat per month, includes one concurrent bot and unlimited meetings, and provides REST API, WebSocket, and a UI dashboard. Self-hosting is based on the Apache 2.0 license, with no license fee. Specific terms for the enterprise edition and activation sprints (a fixed two-week fee) are not yet provided.