LoginSign Up
← Back to all sparks
R

RAGFlow

AI-ML-PLATFORMS
Velocity6.3

Open source RAG engine for deep document understanding

ragflow.io ↗

RAGFlow's nightlies add tenant roles and enterprise hooks while 1.0 waits.

ragmulti-tenantenterprise editionmodel providersopen-source
◆Current state
RAGFlow's November 13 nightlies are maintenance: a minimatch dependency bump, open-source no-op overlays moved into an enterprise stub, and a NATS test fix. The days before assigned roles to tenants in the admin, added a web extension point for the enterprise edition and added Shengsuanyun as a model provider. 1.0.0-rc3 remains the latest release candidate.
◆Where it's heading
Most recent nightlies prepare RAGFlow for multi-tenant and enterprise builds: tenant roles, header slots, parser and web extension hooks. User-facing changes arrive mainly as new model providers and integration fixes.
◆Prediction
Expect 1.0.0 final soon after rc3, and an enterprise edition built on these extension hooks; the entries do not say when.

◆Recent moves

  1. 1d ago

    RAGFlow nightly bumps minimatch in the web app

    A dependency bump in the web front end. No visible change.

    View source ↗
  2. 1d ago

    RAGFlow nightly moves OSS overlays into an enterprise stub

    Open-source placeholder overlays move into a separate enterprise stub module. More of the code split between open-source and enterprise editions.

    View source ↗
  3. 1d ago

    RAGFlow nightly fixes NATS pull capacity tests

    A test change for the NATS task queue's priority lanes. Internal only.

    View source ↗
  4. 1d ago

    RAGFlow nightly adds a web extension point for enterprise builds

    A hook in the web app that the enterprise edition can fill. It continues the series of seams separating open-source and enterprise code.

    View source ↗
  5. 2d ago

    RAGFlow nightly adds Shengsuanyun as a model provider

    Another model provider joins the list, which matters to teams in China that route models through Shengsuanyun. A small broadening of where RAGFlow can get its models.

    View source ↗
  6. 2d ago

    RAGFlow nightly adds user and tenant names to ingest logs

    Ingest logs now show who and which tenant triggered a job. Useful for multi-tenant operators, invisible to everyone else.

    View source ↗