AFFiNE vs RAGFlow
Short answer
AFFiNE: Notion- and Miro-style workspace (docs, whiteboards, databases) with built-in AI, self-hostable with Docker. RAGFlow: RAG engine with deep document parsing (OCR, layout and tables) and cited answers. Both are in our Knowledge base & RAG category.
Choose AFFiNE if you want: cheaper to host (from €35.99/mo vs €69.99/mo) and lighter official recommended (2 GB vs 16 GB RAM).
Choose RAGFlow if you want: fewer license conditions (Permissive vs Open core).
Last updated · computed from each project's official docs and live GitHub data · sources
AFFiNE vs RAGFlow side by side
| AFFiNE | RAGFlow | |
|---|---|---|
| Category | Knowledge base & RAG | Knowledge base & RAG |
| What it is | Notion- and Miro-style workspace (docs, whiteboards, databases) with built-in AI, self-hostable with Docker. | RAG engine with deep document parsing (OCR, layout and tables) and cited answers. |
| License | MIT (Community Edition) with Enterprise Edition license for server code | Apache-2.0 |
| License type | Open core | Permissive |
| License conditions | The README says the Community Edition is free to self-host under the MIT license. The server code under packages/backend carries an AFFiNE Enterprise Edition license that applies to parts not distributed as Community Edition. Self-hosted workspaces get up to 10 seats; more seats need a paid Team Plan or Installable License. | Permissive. Internal and commercial use allowed; keep license and notices. |
| Status | Active | Active |
| GitHub stars | 74.0k | 93.0k |
| Last commit | Oct 27, 2026 | Oct 29, 2026 |
| Latest release | v0.27.4 | v1.0.0-rc2 |
| Minimum hardware | Not published | Not published |
| Recommended hardware | 4 vCPU · 2 GB RAM | 4 vCPU · 16 GB RAM · 50 GB disk |
| RAM measured by us | Not tested yet | Not tested yet |
| GPU | Not required / not stated | Not required / not stated |
| Est. server cost | from €35.99/mo | from €69.99/mo |
| Deploy method | Docker Compose (official images) | Docker Compose |
| Known gotchas | 4 listed | 2 listed |
| Telemetry | Not documented | Not documented |
| Calls home by default | Not documented | Not documented |
| SSO (OIDC / SAML / LDAP) | Free with limits (OIDC free) | Free (OIDC free) |
| Audit log | Not documented | Not documented |
| RBAC | Not documented | Basic free, more in paid plan |
| Details | Full AFFiNE guide → | Full RAGFlow guide → |
Key differences
- License: AFFiNE: MIT (Community Edition) with Enterprise Edition license for server code (Open core). RAGFlow: Apache-2.0 (Permissive).
- Official hardware: AFFiNE: recommended 4 vCPU · 2 GB RAM (no separate minimum published). RAGFlow: recommended 4 vCPU · 16 GB RAM · 50 GB disk (no separate minimum published).
- Est. server cost: AFFiNE: from €35.99/mo. RAGFlow: from €69.99/mo.
- Deploy: AFFiNE: Docker Compose (official images) · multi-container stack. RAGFlow: Docker Compose · multi-container stack.
- GitHub stars: AFFiNE: 74.0k. RAGFlow: 93.0k.
- Last commit: AFFiNE: Oct 27, 2026. RAGFlow: Oct 29, 2026.
- Latest release: AFFiNE: v0.27.4 (Aug 18, 2026). RAGFlow: v1.0.0-rc2 (Oct 16, 2026).
Same for both: Category: Knowledge base & RAG; GPU: not required or not stated; Status: Active.
AFFiNE vs RAGFlow: FAQ
AFFiNE or RAGFlow: which should a small business choose?
Choose AFFiNE if you want: cheaper to host (from €35.99/mo vs €69.99/mo) and lighter official recommended (2 GB vs 16 GB RAM). Choose RAGFlow if you want: fewer license conditions (Permissive vs Open core).
Which is cheaper to self-host, AFFiNE or RAGFlow?
AFFiNE: cheaper to host (from €35.99/mo vs €69.99/mo).
How much hardware do AFFiNE and RAGFlow need?
AFFiNE: recommended 4 vCPU · 2 GB RAM (no separate minimum published). RAGFlow: recommended 4 vCPU · 16 GB RAM · 50 GB disk (no separate minimum published).
What is the license difference between AFFiNE and RAGFlow?
AFFiNE: MIT (Community Edition) with Enterprise Edition license for server code (Open core). The README says the Community Edition is free to self-host under the MIT license. The server code under packages/backend carries an AFFiNE Enterprise Edition license that applies to parts not distributed as Community Edition. Self-hosted workspaces get up to 10 seats; more seats need a paid Team Plan or Installable License. RAGFlow: Apache-2.0 (Permissive). Permissive. Internal and commercial use allowed; keep license and notices. This is a summary, not legal advice.
Is AFFiNE or RAGFlow more actively maintained?
AFFiNE: Active, last commit Oct 27, 2026. RAGFlow: Active, last commit Oct 29, 2026. GitHub data refreshed Oct 29, 2026.
Which is easier to deploy, AFFiNE or RAGFlow?
AFFiNE installs via Docker Compose (official images) (multi-container stack); RAGFlow via Docker Compose (multi-container stack). AFFiNE lists 4 known gotchas and RAGFlow lists 2; see each app page.
Related
- AFFiNE self-hosting guide · RAGFlow self-hosting guide
- All AFFiNE alternatives, ranked
- All RAGFlow alternatives, ranked
- AFFiNE vs AnythingLLM
- AFFiNE vs DocsGPT
- AFFiNE vs FastGPT
- AFFiNE vs kotaemon
- AFFiNE vs MaxKB
- AFFiNE vs Onyx (formerly Danswer)
- AFFiNE vs Open Notebook
- AFFiNE vs PipesHub
- AFFiNE vs PrivateGPT
- AnythingLLM vs RAGFlow
- Coze Studio vs RAGFlow
- Dify vs RAGFlow
- Add a third app in the interactive comparison
Sources
- AFFiNE GitHub repository: github.com/toeverything/AFFiNE
- AFFiNE documentation: docs.affine.pro/self-host-affine
- AFFiNE license: github.com/toeverything/AFFiNE/blob/canary/LICENSE
- AFFiNE hardware requirements: docs.affine.pro/self-host-affine/install/requirements
- AFFiNE deploy instructions: docs.affine.pro/self-host-affine/install/docker-compose-recommended
- RAGFlow GitHub repository: github.com/infiniflow/ragflow
- RAGFlow documentation, RAGFlow hardware requirements, RAGFlow deploy instructions: github.com/infiniflow/ragflow#-get-started
- RAGFlow license: github.com/infiniflow/ragflow/blob/main/LICENSE
- Hetzner Cloud pricing used for cost estimates (read 2026-10-25): www.hetzner.com/cloud/regular-performance
- GitHub REST API (stars, commits, releases): docs.github.com/en/rest
Every value is generated from the same verified data as the app pages. Spotted something wrong? Tell us.