Vexavexa
Book a demo
Comparison

Vexa vs Meeting BaaS

Meeting BaaS gives you a bot API. Vexa gives you the bot, plus the knowledge and agents on top, on your own servers.

Where your
meetings live

Both send a bot to the call. The difference is where recordings end up, and what self-hosting takes.

Where recordings and transcripts live

Vexa, self-hosted

Recordings in your object storage, transcripts in your database.

Meeting BaaS

Meeting BaaS’s cloud for 3 to 30 days, or your S3-compatible storage.

Running it yourself

Vexa, self-hosted

A public repository. Deploy with Docker or Helm; nothing to request.

Meeting BaaS

Private Helm charts: contact your Meeting BaaS representative for access.

The licence

Vexa, self-hosted

Apache-2.0, perpetual and irrevocable: we cannot withdraw it from your copy.

Meeting BaaS

Bot: Elastic License 2.0, which forbids offering it as a managed service.

Which model it runs on

Vexa, self-hosted

A speech-to-text model you choose and deploy, via an OpenAI-compatible endpoint.

Meeting BaaS

Gladia transcription, or your own transcription provider’s key.

Vexa is also available hosted, at the prices below. Meeting BaaS details are from its pricing, self-hosting documentation and bot repository licence, read 3 October 2026.

Feature
comparison

FeatureVexaMeeting BaaS
Licence
Apache-2.0
Elastic License 2.0bot repository
Self-hostable
access on request
Helm chart for Kubernetes
private repository
Google Meet
Microsoft Teams
Zoom
Real-time transcription
Recording + storage
REST API
WebSocket streaming
Webhooks
MCP server
Price
Vexa
Hosted: $0.30 per bot-hour, plus $0.20/hr transcription.
Self-hosted: no licence fee.
Meeting BaaS
From $0.35/hr recording, plus transcription and streaming; meetingbaas.com/pricing, read 3 October 2026.
Knowledge and agents
Knowledge from your meetings
Vexa
A workspace of people, decisions and notes, on your servers (early)
Meeting BaaS
Transcripts by API, searchable through its MCP; you build the knowledge.
Agents on that knowledge
Vexa
Self-hosted: an agent drafts the write-up; a person approves it (early)
Meeting BaaS
Speaking AI bots, an MCP server for Claude, and optional summaries.

Knowledge and agents: from Meeting BaaS’s notetaker guide, MCP server, speaking bots and transcription docs, read 4 October 2026.

Run it on your own servers

Deploy it yourself or start on our hosted API: your code calls the same API either way.

No credit card required

Frequently asked questions

Can both Vexa and Meeting BaaS be self-hosted?

Yes. Vexa’s code is public under Apache-2.0, so you can deploy it today. Meeting BaaS documents a self-hosted version on Kubernetes; its docs say to contact your Meeting BaaS representative for access to the Helm charts.

What licence is each under?

Vexa is under Apache-2.0. Meeting BaaS’s public bot repository is under the Elastic License 2.0, which does not allow offering it to others as a hosted or managed service. Its self-hosting docs name no licence for the full platform.

Do both support the same meeting platforms?

Yes. Both send bots to Google Meet, Microsoft Teams and Zoom, stream in real time over WebSocket and send webhooks. Both also offer an MCP server.