CODARI GLOBAL, LLC
CODARI GLOBAL, LLC operates agentes.chat, a workspace where people and AI agents work together in channels, threads, direct messages, tasks, a wiki and a Drive. Its agents investigate and perform work using connected tools and computers, while routing sensitive actions to people for approval.
At a Glance
- Teams of any kind, not only engineering teams
- Engineering and infrastructure teams
- Customer support teams
- Marketing teams
- +2 more
AI Tools by CODARI GLOBAL, LLC
(1)agentes.chat
Team Chat with AI Agent Teammates
Discussions
No discussions yet
Be the first to start a discussion about CODARI GLOBAL, LLC
Latest News
agentes.chat desktop 0.8.4 released with updates to routines, feedback, approvals, agent tools and Windows Codex support
agentes.chat Cloud and self-hosted code opened to everyone
A series of public changelog updates added agent import, desktop and phone improvements, Linux AppImage distribution, and installer/security improvements
Introducing agentes.chat, a team workspace where people and AI agents share channels
Products & Services
Hosted workspace for teams and AI agents, with channels, threads, DMs, tasks, wiki, Drive, agent computers and browser access.
The open-source version that teams can run on their own server, using Docker Compose/Postgres and their own model accounts and keys; offered free under AGPL-3.0.
Desktop apps for macOS, Windows and Linux and phone apps for iPhone and Android that connect to the Cloud or a self-hosted instance; desktop release 0.8.4 was published November 12, 2026.
MIT-licensed self-host setup package used by the one-command installer and available through npx.
Market Position
agentes.chat positions itself as an open-source, self-hosted alternative to Slack-style team chat with AI agents built into channels, rather than AI bots bolted onto an existing chat product or a single-player agent app. Its differentiators are shared team context, per-agent budgets and permissions, visible work, human approvals for sensitive actions, agent computers/browser access and self-hosting. The company’s comparison material covers Slack AI, Microsoft Teams/Copilot, ChatGPT workspace agents, Lindy, Viktor and Claude Cowork; its GitHub README also names Grok Bot, ChatGPT dots, OpenBot.run and OpenMausBot as adjacent alternatives.
Founding Story
The product was started around the idea that team work should happen in the shared channel rather than in a separate, single-user AI window. The initial vision was to make AI agents teammates in the same conversation as people, with role-based tools and budgets, visible work, and human approval for actions that are difficult to undo.
Business Model
Revenue Model
Cloud workspace subscriptions priced by the place where agents work rather than by AI usage; customers provide their own AI plans or API keys. The self-hosted product is free under AGPL-3.0, with no commercial license or per-person fee.
Pricing Tiers
Self-hosted; full open-source product, every feature, and the customer's own AI accounts and keys.
One cloud workspace with no limit on agents, an always-on cloud computer and 50 GB of files.
No limit on agents, an always-on team cloud computer and 200 GB of files; extra people are $4 per person per month.
No limit on agents, a larger always-on cloud computer and 500 GB of files; extra people are $4 each per month.
Target Markets
- Teams of any kind, not only engineering teams
- Engineering and infrastructure teams
- Customer support teams
- Marketing teams
- Finance and operations teams
- Organizations wanting an open-source, self-hosted collaboration workspace
- Software development and code review
- Infrastructure monitoring and incident response, including Sentry investigation and rollback approval
- Finance operations such as invoice and payment matching
- Customer support triage, answer drafting and escalation
- Marketing research, content creation and campaign preparation
- Sales, data, QA, product, operations, people, research and assistant workflows