Buzz

Learning Objectives

• Understand what makes Buzz an AI-native team collaboration platform rather than a bolt-on chatbot
• Identify when a decentralized, self-hostable, agent-first workspace is the right fit
• Evaluate Buzz against Slack and other agent-collaboration tools

What Is Buzz?

Buzz is an open-source collaboration platform from Block — Jack Dorsey's company — that puts human teammates and AI agents in the same conversations. Launched on July 21, 2026, it combines the team-chat experience of Slack with the code collaboration of a Git host, and treats AI agents as first-class participants in channels rather than as a separate assistant panel bolted on the side.

Dorsey framed the launch around four properties: Buzz is model-agnostic, decentralized, self-sovereign, and open source. In practice that means teams choose which AI models power their agents, the underlying protocol is not owned by a single vendor, users keep control of their own data, and the full source is published for anyone to inspect, customize, or extend.

Pricing

Buzz is free and Apache-2.0 licensed at launch. Teams can run their own instance on their own infrastructure for full control, or create a community on Block-provided hosting at buzz.xyz to skip the setup. Because the platform is open source and decentralized, there is no per-seat subscription gate on the core product the way there is with proprietary team-chat tools.

Core Features

Humans and Agents in One Workspace

Buzz's defining idea is that AI agents live inside the same channels, threads, and direct messages as people. Instead of copy-pasting between a chat app and a separate agent tool, a team can @-mention an agent in a thread, hand it a task, and watch it work alongside human replies. Because Buzz is model-agnostic, each team wires in whichever models it prefers rather than being locked to one provider's assistant.

Team Chat Plus Git Hosting

Buzz bundles the collaboration primitives most teams already expect — channels, threads, direct messaging, voice, and media sharing — with built-in Git hosting. That combination targets the daily context-switch between a chat tool (Slack) and a code host (GitHub): pull requests, repositories, and conversations sit on one platform, so agents and humans can act on code without leaving the workspace.

Automated Workflows

Beyond conversation, Buzz supports automated workflows so recurring tasks — triage, notifications, routine code or ops actions — can be handled by agents on triggers rather than by hand. This is the "agents as teammates" thesis applied to process: the workspace is where work gets orchestrated, not just discussed.

Open, Decentralized, and Self-Hostable

The entire app is published on GitHub under Apache-2.0, and the Nostr foundation means an organization can run Buzz on its own infrastructure with its own identity layer. For security- or sovereignty-conscious teams, that removes the single-vendor dependency that comes with proprietary SaaS chat.

Strengths

• Agent-native by design: AI agents are first-class channel participants, not a bolted-on assistant
• Model-agnostic: teams choose their own models rather than being tied to one vendor's assistant
• Chat + Git in one place: collapses the Slack-plus-GitHub context switch for engineering teams
• Open source (Apache-2.0): full source on GitHub to inspect, fork, and extend
• Decentralized and self-sovereign: Nostr-based, self-hostable, user-controlled data and identity
• Free at launch: no per-seat gate on the core product; free desktop app for macOS, Windows, and Linux

Limitations & Considerations

• Early-stage: Block describes Buzz as in its early stages — expect rough edges, gaps, and rapid change versus a mature tool like Slack
• Ecosystem maturity: Slack's decade of third-party integrations, apps, and enterprise controls is not something a brand-new platform matches on day one
• Nostr learning curve: the decentralized, self-sovereign model is powerful but less familiar than a standard hosted SaaS account for non-technical teams
• Migration cost: moving an existing organization's history, integrations, and habits off an incumbent chat tool is real work, regardless of the destination
• Competitive field: other agent-first workspaces are emerging (for example, the open-source Centaur from Paradigm's CTO), and the category is moving quickly

Best Use Cases

When to choose alternatives:

• Mature enterprise chat with deep integrations → Slack or Microsoft Teams
• Pure code hosting and review → GitHub or GitLab
• A single-vendor assistant inside existing tools → Microsoft Copilot or a vendor-native agent

Getting Started

1. Visit buzz.xyz and create a community on Block-provided hosting, or download the free desktop app for macOS, Windows, or Linux
2. For full control, clone the source from github.com/block/buzz and self-host your own instance
3. Set up channels, threads, and direct messages the way you would in any team-chat tool
4. Connect your preferred AI models and invite agents into channels — Buzz is model-agnostic, so wire in the providers your team already uses
5. Move a repository into Buzz's built-in Git hosting and try a workflow where an agent and a human collaborate on the same thread
6. Start small with one team or project before migrating an entire organization off an incumbent tool

Key Takeaways

• Buzz is Block's open-source, Nostr-based collaboration platform where humans and AI agents work together in the same channels — launched July 21, 2026 by Jack Dorsey's Block
• It combines Slack-style team chat with built-in Git hosting and automated workflows, targeting the daily context-switch between a chat app and a code host
• The platform is model-agnostic, decentralized, self-sovereign, and Apache-2.0 licensed, with a free desktop app for macOS, Windows, and Linux that teams can self-host or run on Block-provided hosting
• It is deliberately agent-native — AI agents are first-class participants, not a bolted-on assistant — which is the whole reason to choose it over a retrofitted incumbent
• Buzz is early-stage: promising and free, but not yet a match for Slack's ecosystem maturity and enterprise controls