Rasa
Build sophisticated AI assistants with no-code customization and seamless integration.. [Contact for Pricing]
Last verified:
What is Rasa?
Rasa is an enterprise developer platform for building, deploying, and managing conversational AI agents across every channel. It extends large language models (LLMs) with business logic to create reliable AI agents that handle millions of conversations while giving teams full control over behavior and performance. The platform uses CALM (Conversational AI with Language Models) for dialogue management, enabling a shift from traditional intent-driven systems to LLM-based agents that adhere to business logic and guard against prompt injection and hallucination.
The Rasa Platform includes a pro-code interface called Rasa Pro for developers and a no-code interface called Rasa Studio for business users. Key features include patented dialogue management, on-prem/private cloud deployment, multi-agent orchestration, voice/IVR support with low-latency processing, real-time observability, built-in security against hallucination and jailbreaking, enterprise search, contextual response rephraser, custom actions server via Rasa SDK, channel connectors, Kubernetes deployment support, end-to-end testing, PII data management, and multi-LLM management.
Rasa is designed for enterprise teams in regulated industries who need data privacy, security, and compliance. It serves customer support teams building customer support agents, IT/HR teams creating internal helpdesk agents, organizations needing search agents, and companies deploying voice agents. Top enterprises like Providence Health, N26, ERGO Group, Albert Heijn, and nib Group trust Rasa for mission-critical conversational AI applications.
The platform supports omnichannel deployment across web, mobile, messaging apps, and voice channels. Teams can build conversational voice or chat agents using the open developer framework with a coding copilot, deploy on their own infrastructure while Rasa orchestrates conversations and external agents, and refine agent performance in Rasa Studio with real-time monitoring and conversation analysis tools.
Rasa pricing
Pricing model: Freemium
Rasa offers three subscription tiers. The Free Developer Edition costs $0/month and includes a free Rasa Pro license usable locally or in production, one bot per company, up to 1000 external conversations/month (or 100 internal conversations/month for employee agents), CALM, Enterprise Search, E2E testing, and free community support via forum. The Growth plan is for teams with under 500,000 conversations annually and includes simple deployment, no-code UI for faster time-to-market, Basic Support with web-based support through online portal, break & fix support, and best effort response times during normal business hours. The Enterprise plan requires contacting sales and includes full access to Rasa Platform, Premium Support with 24/7/365 enhanced response times, Customer Success Manager and Customer Success Engineer access, success planning, best practice guidance, business reviews, large-scale deployment capabilities, increased automation rates, and pre-built enterprise security features. Rasa Studio (no-code UI) includes everything in Rasa Pro plus no-code AI assistant flow builder, testing panel, content/response management system, conversation view for analysis, Open Prompt editing, Single-Sign-On (SSO), and Role-based Access Control (RBAC). Deployment modes include Self-Managed (On-prem or Private Cloud) and Managed Service. IVR Connector to AudioCodes VoiceAI Connect available through additional purchase.
Rasa pros
- Complete data privacy with on-premise and private cloud deployment
- Patented dialogue management system for enterprise AI agents
- CALM prevents LLM hallucination and prompt injection by design
- Full control over prompts, policies, and underlying codebase - no black boxes
- Multi-agent orchestration with shared conversational state and memory
- Built-in voice/IVR support with low-latency streaming and turn-taking
- Real-time observability with conversation tracing and decision path inspection
- Open framework with no vendor lock-in - choose any LLM and infrastructure
- Scales to millions of conversations with proven enterprise deployments
- Rasa Studio provides no-code interface for business users
- Rasa Pro offers pro-code interface with coding copilot for developers
- End-to-end testing and PII data management built-in
- Kubernetes deployment support through Helm charts
- Multi-LLM management and LLM fine-tuning recipe included
- 50% cost reduction in operational expenses across enterprise deployments
- 59% goal completion rate per conversation reported by customers
- 4.4 star customer satisfaction consistently maintained
- Secrets management powered by Vault and dependency vulnerability protection
Rasa cons
- Steeper learning curve compared to no-code conversational AI platforms
- Requires ML and engineering expertise for effective implementation
- Self-hosting requires infrastructure management and DevOps capabilities
- Free Developer Edition limited to 1000 external conversations per month
- Rasa Studio does not support all functionality of Rasa Pro
- Enterprise plan pricing not publicly disclosed - requires sales contact
- Only one bot per company in free Developer Edition
- Analytics dependencies on Kafka event broker limit flexibility
Frequently asked questions about Rasa
What is Rasa and what does it do?
Rasa is an enterprise developer platform for building, deploying, and managing conversational AI agents across every channel. It extends large language models (LLMs) with your business logic to create reliable AI agents that handle millions of conversations. Rasa uses CALM (Conversational AI with Language Models) for dialogue management, which enables teams to shift from traditional intent-driven systems to LLM-based agents that adhere to business logic and guard against prompt injection and hallucination. The platform gives your team full control over agent behavior and performance.
Is Rasa free to use?
Yes, Rasa offers a Free Developer Edition that costs $0/month. This includes a free Rasa Pro license usable locally or in production, one bot per company, up to 1000 external conversations per month (or 100 internal conversations/month for internal employee agents), CALM dialogue management, Enterprise Search, E2E testing, and free community support via their forum. The Developer Edition can be used both locally and in production.
Can I deploy Rasa on my own infrastructure?
Yes, Rasa supports on-prem/private cloud deployment. You can deploy Rasa on your own infrastructure, private cloud, or fully offline environments—including air-gapped deployments when required. Organizations retain full control over their data, security, and model providers. This is built with regulated industries in mind, enabling teams to meet strict compliance and governance requirements without sacrificing flexibility in infrastructure, integrations, or AI model choice.
What is CALM in Rasa?
CALM stands for Conversational AI with Language Models. It is Rasa's approach to dialogue management that enables you to shift from traditional intent-driven systems to LLM-based agents. CALM allows you to build more robust agents that are easier to create and maintain, adhering to business logic and guarding against prompt injection and hallucination by design. CALM is resistant to hallucination, prompt injection, and jailbreaking, and you can run it fully on-premise with no calls to external LLMs.
What is the difference between Rasa Pro and Rasa Studio?
Rasa Pro is the pro-code generative AI native conversational AI framework for developers to flexibly build, integrate, monitor, and deploy AI Assistants. It includes CALM dialogue management, language-agnostic NLU, enterprise search, contextual response rephraser, custom actions server, channel connectors, Kubernetes deployment support, end-to-end testing, PII data management, and more. Rasa Studio is the no-code user interface for business users that includes everything in Rasa Pro plus a no-code AI assistant flow builder, testing panel, content and response management system, conversation view for analysis, Open Prompt editing, Single-Sign-On (SSO), and Role-based Access Control (RBAC).
Does Rasa support voice and IVR?
Yes, Rasa supports real-time voice interactions designed for natural, human conversation with low-latency processing enabling streaming speech recognition, fast turn-taking, and responsive dialogue flow across phone and voice channels. The system supports core voice primitives such as barge-in, turn-taking, and background noise handling, allowing AI agents to manage interruptions, respond quickly, and sustain fluid back-and-forth conversations. Rasa also includes a Voice Gateway with built-in turn-taking, timeouts, and latency control, and offers an IVR Connector to AudioCodes VoiceAI Connect available through additional purchase.
Which companies use Rasa?
Top enterprises trust Rasa including Providence Health (Digital Innovation Team), N26, ERGO Group AG, Albert Heijn (largest supermarket chain in the Netherlands), nib Group, Orange, Groupe IMA, eUprava (government services), and Eddy Travels. These companies use Rasa for customer support agents, internal helpdesk agents, patient-facing assistants, government services, and voice agents across regulated industries like healthcare, banking, retail, and government.
What channels does Rasa support?
Rasa supports omnichannel deployment across every channel. This includes Text & Messaging with built-in support for web, mobile, and messaging apps with natural, fluid conversations. Rasa provides channel connectors via REST and WebSocket, SocketIO channel for real-time websockets, RestInput and CallbackInput for custom integrations, and a Rasa Chat Widget that can be incorporated into existing webpages. The platform also supports voice/IVR channels and can integrate with Slack, Facebook, and other messaging platforms.
How does Rasa handle security and compliance?
Rasa provides built-in security with CALM being resistant to hallucination, prompt injection, and jailbreaking by design. You can run it fully on-premise with no calls to external LLMs, ensuring complete data privacy. The platform includes PII data management, secrets management powered by Vault, dependency vulnerability protection, and pre-built enterprise security features in the Enterprise plan. On-prem/private cloud deployment allows organizations to retain full control over their data, security, and model providers, making it suitable for regulated industries with strict compliance requirements.
What observability and monitoring tools does Rasa provide?
Rasa provides real-time observability and control with built-in tools to monitor, evaluate, and improve AI agent behavior in production. Teams can trace conversations, inspect decision paths, and analyze workflow performance to quickly diagnose issues. The platform includes transparent logs, evaluation tools, feedback loops, conversation view in Rasa Studio to analyze conversations, and observability via OpenTelemetry. With these tools, organizations maintain operational oversight while continuously optimizing agent reliability, safety, and business impact.