Matagi

Matagi is a no-code platform for building and running AI agents. It supplies everything an agent needs to operate — servers, storage, memory, inboxes, tool access, and model calls — and provisions each piece automatically as the agent is described, so no configuration files or API credentials have to be handled manually. The platform is designed to work alongside existing AI tools such as Claude, OpenAI, and Cursor, positioning itself as the execution layer those assistants lack rather than a replacement for them. Agents can be connected to more than 3,000 tools, including Gmail, Outlook, Slack, Microsoft Teams, Google Calendar, HubSpot, Salesforce, Notion, and Stripe. Common use cases span revenue and back-office work, from lead enrichment and CRM hygiene to inbox triage, meeting follow-up, invoice retrieval, and client reporting.

Features

• Automatic Provisioning: Spins up the servers, storage, memory, inboxes, and model access an agent needs as it is being built, with no config files to touch.
• No-Code Agent Creation: Agents are described in plain language, and the platform handles how they are assembled and run.
• 3,000+ Tool Integrations: Connects agents to common business tools including Gmail, Outlook, Slack, Teams, Google Calendar, HubSpot, Salesforce, Notion, and Stripe.
• Works With Existing AI Tools: Plugs into Claude, OpenAI, and Cursor so teams keep the assistant they already use while Matagi runs the resulting agent.
• API and MCP Access: Every plan, including the free trial, includes API and MCP integration for reaching agents from other systems.
• Prebuilt Agent Templates: Ready-made starting points such as inbox triage, lead enrichment, CRM hygiene, competitor monitoring, support triage, and client reporting.

Pricing

Matagi offers four plans, with agent, model, and infrastructure usage billed separately:

• Free trial: $0 for 7 days. One agent project, up to 5 resources per custom agent, and API and MCP integration, with no card required.
• Builder: $49/month. For solo builders shipping their first agents, with 5 agent projects, one seat, up to 5 resources per agent, and community support.
• Team: $249/month. For teams running agents in production, with 50 agent projects, unlimited seats, unlimited resources per agent, and priority support over Slack.
• Enterprise: Contact for details. Custom agent project limits, unlimited seats, and dedicated support for organizations with scale and compliance requirements.

Matagi states that agent runtime, model calls, and infrastructure are passed through at cost with no margin added. Teams that already hold model credits can connect their own provider keys so LLM usage runs on their own accounts.

Pros

• No Infrastructure Setup: Servers, credentials, and integrations are provisioned automatically, removing the engineering work usually needed to put an agent into production.
• Broad Tool Coverage: With more than 3,000 integrations available, most common sales, support, and finance stacks are already supported.
• Transparent Usage Billing: Runtime, model, and infrastructure costs are billed at cost rather than marked up on top of the subscription.
• Low Cost to Start: A seven-day free trial with no card and a $49 monthly plan keep initial testing inexpensive.
• Model Flexibility: Teams can supply their own provider keys instead of purchasing model usage through the platform.
• Extends Tools Teams Already Use: The platform complements Claude, OpenAI, and Cursor rather than requiring a switch to a new assistant.

Cons

• No Public Reviews Yet: The product has no listings on major review platforms, so there is little independent feedback for buyers to weigh.
• Unpredictable Total Cost: Subscription prices are fixed, but usage-based charges for runtime and model calls make the monthly total harder to forecast.
• Resource Caps on Lower Plans: The trial and Builder plan limit each custom agent to five resources, which may constrain more complex builds.
• General-Purpose Rather Than Sales-Specific: Sales teams design their own agents instead of adopting prebuilt outbound workflows tuned for prospecting.
• Results Depend on Specification: Because agents are defined by the user, output quality tracks how clearly the workflow is described, which favors teams comfortable with automation.