One plan for every step of your journey
From POC on Free to team deployment on Scale. No markup on your LLM keys, cancel anytime.
One license for your clients, plans to evaluate
The main channel remains Cloud, support and dedicated deployments. Free, Pro and Scale plans exist to evaluate the platform self-service before deploying.
Deploy IgnitionRAG for your clients with a firm license
Dedicated deployment, guided onboarding, nested firm → client multi-tenancy, priority support, and commercial terms tailored to your portfolio.
Pro
- 15 Collections
- 2,000 Documents
- 20 Workflows
- 5,000 Runs/month
- 5 MCP Servers
- All triggers
- API + SDK access
- 1 User (solo)
Scale
- 50 Collections
- 10,000 Documents
- 100 Workflows
- 50,000 Runs/month
- 50 MCP Servers
- All triggers
- API + SDK access
- 15 Members
- Priority support
Enterprise License
Firm license
Self-hosted or hosted by us
- Hands-on onboarding
- Nested multi-tenant (firm → your clients)
- Dedicated deployment
- Unlimited everything
- Your own LLM keys
- Guaranteed SLA
- Dedicated support
- Custom domain
- SSO / SAML
Questions? Answers.
Everything you need to know.
What is IgnitionRAG?
IgnitionRAG is the AI infrastructure layer that turns your enterprise knowledge into production-ready AI assistants. It bundles the whole RAG stack (ingestion, retrieval, agents, evaluation, observability, access control) into one platform, so your team ships in weeks.
How fast can we launch our first assistant?
Most teams go from documents to a working, production-grade assistant in weeks, and a first proof-of-value can run in days. You spend your time on the use case and adoption, not on infrastructure.
Do we have to rebuild ingestion, retrieval and agents ourselves?
No. Multimodal ingestion, hybrid search with reranking, agents with tools, evaluation and observability come as one platform. You don't assemble five frameworks and glue them together.
Is it production-ready for enterprise?
Yes. Evaluation (Precision, Recall, MRR and LLM-judge), full observability over traces, latency, tokens and cost, plus governance and role-based access control are part of the platform, not add-ons you build later.
How do our engineers integrate it?
Through a REST API with OpenAPI, official TypeScript and Python SDKs, and a native MCP server (32 tools) for Claude, Cursor and your IDE. Everything the platform does is reachable from code, never locked behind our UI.
Can business teams run it without engineers?
Yes. Business stakeholders manage imports, user feedback, costs and ROI from a dashboard, and answers ship through an embeddable widget or your existing channels. No engineer in the loop for day-to-day operation.
Which LLMs can we use, and how does pricing work?
IgnitionRAG is model-agnostic and BYOK (bring your own key): OpenAI, Anthropic, Mistral, Azure OpenAI, or any OpenAI-compatible endpoint. You connect your own keys and pay the providers directly, with zero markup on your tokens.
Where is our data hosted, is it compliant, and can we self-host?
Hosting is in France/EU with GDPR compliance, audit logs and right-to-erasure built in. With BYOK your data isn't routed through our accounts, and you can self-host via Docker or run a dedicated deployment with SSO/SAML, SLA and onboarding (Enterprise).
See your first enterprise AI assistant come to life
Book a 30-minute demo. We'll map your documents, your use case and the fastest path from enterprise knowledge to a production assistant.