Read AI

You've sat through the meeting, nodded at the decisions, and two days later nobody can agree on who owned the follow-up. The notes are scattered across someone's Google Doc, a Slack thread and a half-remembered Zoom chat. Read AI is built for that exact gap. Its pitch on read.ai is blunt: "Meetings In, Decisions Out." The idea is that every call should end with recorded outcomes, not just a recording.

In practice, Read AI joins your calls on Google Meet, Zoom and Microsoft Teams and produces recaps, action items, highlights, a transcript and video. Before the meeting, it pulls context from your calendar, email, past meetings and CRM so you know "what was promised last time and what's still open." After the meeting, action items get routed into the tools you already use, such as HubSpot, Salesforce, Notion, Asana, Atlassian and Slack. The part that sets it apart from a basic notetaker is Ask Read, which answers questions like "who owns this?" across meetings, emails and messages, with citations. It's also now available as a Claude Connector, a ChatGPT App and an MCP Server, so you can query your meeting history from inside the AI tools you already use.

The free tier is a real trial rather than a teaser: 5 meeting transcripts a month, unlimited search and apps for Windows, macOS, Android and iOS, with no credit card. Paid plans start at $22.50/mo, and annual billing knocks 25% off. On privacy, Read AI says it's SOC 2 Type 2 certified, GDPR and HIPAA compliant, and doesn't train on your data by default. The catch is sprawl. Between meetings, email, messaging, workspaces and a "digital twin" assistant called Ada, the product is trying to do a lot, and some of those pieces are described in far less detail than the core notetaker.