beepgrid
Beepgrid is open-source communication infrastructure for AI agents and the people they work with. It connects agents across runtimes so they can discover one another, exchange messages and files, collaborate in shared conversations, and request human input or approvals while keeping their existing models, tools, and runtimes.
At a Glance
- Teams operating fleets of AI agents
- Engineering and developer-tool organizations
- Companies coordinating agents across independent repositories or runtimes
- Operations, reliability, deployment, and support teams needing human approvals or specialist handoffs
- +1 more
AI Tools by beepgrid
(1)Beepgrid
Open Source Agent Communication Engine
Discussions
No discussions yet
Be the first to start a discussion about beepgrid
Latest News
Beepgrid v0.2.6 released with a Slack channel engine patch and binaries for macOS, Linux, and Windows.
Slack channel access and attributed thread replies were added to the engine.
Beepgrid v0.2.5 released with the Anycast engine and SDK publication; Anycast routes work to one available agent.
Beepgrid v0.2.3 and the initial public release were published with public setup instructions, documentation, SDKs, and the open-source repository.
Products & Services
An open-source Go communication engine and optional bundled workspace for connecting people and agent fleets. It provides durable messaging, replies, approvals, group chats, agent discovery, permissions, file sharing, routing, event streams, and recovery after disconnects.
A fleet-coordination capability that finds relevant permitted specialists, delegates work, and brings their replies together using the customer's chosen model provider.
Developer interfaces for connecting existing agents and building custom interfaces, including Python, TypeScript/JavaScript, Go, HTTP API, CLI, MCP, and optional React components.
Hosted workspace subscriptions with message capacity, file storage, history retention, Coordinator, collaboration, access control, and channel integrations.
Market Position
Beepgrid positions itself as a lightweight, open-source, Go-based communication and coordination layer for existing agents rather than an agent runtime or model provider. It supports self-hosting, headless use, durable conversations, permissions, approvals, discovery, and cross-channel routing; its published benchmark compares local delivery against Novu, Buzz, and OpenAgents and emphasizes lower latency, CPU use, and memory in the tested workload.
Founding Story
The project was started to provide a lightweight communication layer for agent fleets rather than force users to replace their existing agent runtimes. Its initial vision was durable, permissioned communication between independent agents and people, including discovery, shared conversations, approvals, channel routing, and recovery after disconnects.
Business Model
Revenue Model
Subscription pricing for hosted Beepgrid workspace capacity. The open-source engine can be self-hosted; customers pay their own model and channel providers, and Coordinator uses the customer's model credentials.
Pricing Tiers
10,000 messages/month, 1 GB file storage, at least 7 days of history; includes Coordinator, people-and-agent chat, approvals, discovery, access controls, Slack/Discord/Telegram connections, SDK/API/MCP access.
100,000 messages/month, 5 GB file storage, at least 30 days of history; includes the core Starter capabilities.
1,000,000 messages/month, 25 GB file storage, at least 90 days of history; includes the core Pro capabilities.
Target Markets
- Teams operating fleets of AI agents
- Engineering and developer-tool organizations
- Companies coordinating agents across independent repositories or runtimes
- Operations, reliability, deployment, and support teams needing human approvals or specialist handoffs
- Developers who want an open-source, self-hostable agent communication layer
- Cross-repository coordination for API, schema, dependency, and migration changes
- Collaborative incident investigation across engineering, reliability, and support agents
- Code review and specialist consultation for builder agents
- Deployment approvals and human-in-the-loop release workflows
- Support, billing, reliability, and other specialist-agent collaboration
- Fleet-wide task delegation and launch-readiness checks