- vendor
- Ragie
- whatItIs
- Full-stack managed RAG pipeline that handles document ingestion, parsing, chunking, vector indexing, keyword indexing, summary indexing, entity extraction, hybrid retrieval, and reranking behind a single API. Targets agent-native use cases requiring fresh, entity-aware context.
- hosting
- both
- pricingModel
- free tier (1,000 docs); Starter $100/mo (10,000 pages); Pro $500/mo (60,000 pages); Enterprise custom; connectors $250/mo each
- modelProviders
- agnostic
- stateModel
- persistent indexed corpus per partition; multi-index (vector + keyword + summary)
- toolModel
- REST API + Python/TypeScript SDKs + built-in source connectors (Google Drive, Notion, Confluence, S3)
- includesBrowser
- false
- includesMemory
- false
- longRunning
- continuous ingestion pipelines
- hitlSupport
- false
- observability
- retrieval quality metrics + connector sync logs
- maturity
- ga
- launched
- 2024
- notes
- SOC 2, GDPR, HIPAA. Multimodal input: text, PDF, images, audio, video, PowerPoint. Distinct from Vectara in its full multi-index approach and entity extraction; distinct from Unstructured in that it also owns retrieval, not just parsing.
