RAGFlow 1.0.0: agent platform reaches stable release
The stable release of RAGFlow 1.0.0 came out on 06.11.2026. It is a platform for building AI agents with a knowledge base, which the authors describe as an agent-native database — a RAG engine that has grown into context infrastructure.
What's new
Since the last review, two releases have shipped: 0.27.3 and 1.0.0. The main changes:
- Agentic RAG ported to Go — the agentic retrieval core has been rewritten in the language RAGFlow itself is written in, which simplifies deployment and maintenance.
- New search provider AnySearch — agents can use web search as a context source.
- Wally support as a model provider — another way to connect a model to the platform.
- Virtualization of message lists in chat — the chat interface has become lighter with long conversations.
- User administration — search, status filtering, sorting and pagination in the CLI.
- Pagination fix in MCP — the rerank_candidates_count parameter is now passed correctly, and retrieval works with results beyond 64.
What the platform does
RAGFlow is a RAG engine with agentic capabilities: knowledge base, graphs and state persistence, handling of documents, images, scans and web pages. It deploys via Docker or from source, and installs via npm install. License — Apache-2.0.
Your own model instead of the cloud
You can connect your own model via Ollama, vLLM or an OpenAI-compatible API. The documentation states directly: "You can use Ollama or vLLM to deploy local models", and for models with an OpenAI-compatible API there is a separate setting on the Model providers page. Local deployment — via Docker or running from source.
How it differs from its neighbors
Among its neighbors in the visual agent builder niche, RAGFlow stands out for its stated handling of images and documents — n8n, Dify and Langflow don't mention this in their documentation. At the same time, a visual editor, tracing and debugging, custom tools and functions, and MCP support are not stated in RAGFlow's documentation, although most of its neighbors have them. Graphs, state persistence and a self-hosted server with Docker are stated for all four platforms.
| RAGFlow | n8n | Dify | Langflow | |
|---|---|---|---|---|
| Type | builder | builder | builder | builder |
| Language | Go | TypeScript | Python | Python |
| License | Apache-2.0 | Sustainable Use Licen… | Apache 2.0 with add. terms… | MIT |
| GitHub stars | 93 703 | 209 583 | 159 392 | 155 815 |
| Releases in 90 days | 5 | 58 | 4 | 15 |
| Multiple agents in a chain | — | — | — | yes |
| Custom tools and functions | — | — | yes | yes |
| MCP support | — | — | yes | yes |
| Knowledge base (RAG) | yes | — | yes | yes |
| Sandboxed code execution | — | — | yes | — |
| Human approval | — | yes | — | — |
| Graphs and state persistence | yes | yes | yes | yes |
| Visual editor | — | yes | yes | yes |
| Tracing and debugging | — | yes | yes | yes |
| Voice and speech | — | — | yes | — |
| Self-hosted server and Docker | yes | yes | yes | yes |
| Images and documents | yes | — | — | — |
| Own model: Ollama | yes | — | — | yes |
| Own model: vLLM | yes | — | — | — |
| Own model: LM Studio | — | yes | — | — |
| Own model: OpenAI-compat. API | yes | yes | yes | — |
Cross-checked against the projects' READMEs and documentation as of 07.11.2026: "yes" means the capability is stated there, a dash means it is not mentioned (which does not mean the capability is absent). Stars and releases — per GitHub data.
Worth updating for those who already use RAGFlow and have been waiting for a stable release with the Go-ported core. Worth a look for those building agents with a focus on document and image processing and who want to connect their own models via Ollama or vLLM. You can follow new releases in the CPU3D news feed.
