Skip to main content
Rasa is the code-first interface for building conversational AI agents whose task execution must follow explicit business logic. Rasa manages dialogue with CALM (Conversational AI with Language Models). Large language models (LLMs) interpret what a user wants, while flows define the steps the agent can take to complete a task. This separation keeps business rules in version-controlled files that teams can review and test. This section documents Rasa and its developer features. Product names, packages, and commands may use Rasa Pro when they refer to the licensed distribution.

Developer Edition

Build and run agents for up to 1,000 conversations per month with a free Rasa Developer Edition license.

​
What you can build with Rasa

  • Explicit business logic: Define tasks as flows with reviewable steps, conditions, actions, and responses.
  • Conversation handling: Use conversation patterns for corrections, topic changes, clarifications, and other behavior that occurs around a flow.
  • Backend and agent integrations: Connect APIs through Model Context Protocol (MCP) servers and connect other agents through the Agent-to-Agent (A2A) protocol.
  • Custom components: Extend the framework and inspect its source code when you need behavior beyond the built-in components.
  • Testing and observability: Evaluate conversations before release, then use analytics and tracing to investigate production behavior.
  • Deployment control: Run Rasa on your infrastructure and choose hosted or self-managed language models according to your security requirements.

​
Who Rasa is for

Rasa is for developers who need to control agent behavior, integrations, testing, and deployment through code. Teams can use it with Rasa Studio, the no-code interface for conversational AI practitioners. Rasa Developer Edition supports local and production use for up to 1,000 external conversations per month or 100 internal employee conversations per month.

​
Start building