Repolode
Submit
OpenCSGs

OpenCSGs/csghub

Open-source, self-hostable platform for managing LLMs, datasets, spaces and code, positioned as a private, on-premise alternative to Hugging Face.

Open-source alternative to:Hugging Face

View on GitHubWebsite
Stars
4,080
Forks
509
Open issues
23
Last commit
3 days ago

What is it?

CSGHub is an open-source platform for managing large language model assets. It lets teams upload, download, store, verify and distribute LLMs such as DeepSeek and Llama, along with datasets, spaces and code, through a web interface, the git command line, a natural language chatbot, or the CSGHub SDK.

The platform is built from microservice submodules and exposes standardized OpenAPIs so it can be integrated with existing systems. It supports Hugging Face compatible Python SDK usage, and can be deployed on-premise for offline, controlled environments, with Docker Compose and Helm Chart installation options documented.

It is intended as a private, on-premise counterpart to Hugging Face, covering asset lifecycle management, access control and data processing for LLM workflows.

Key features

  • 01Unified management of LLMs, datasets, spaces and code
  • 02Upload, download and distribution via web, git CLI, chatbot or SDK
  • 03Hugging Face compatible Python SDK usage
  • 04Standardized OpenAPIs and microservice architecture for integration
  • 05On-premises deployment for offline operation
  • 06Enterprise-level security and access control

Use cases

  • Hosting an internal private model registry for an organization that cannot use public cloud hubs
  • Managing LLM versions, datasets and deployed applications in one place
  • Integrating LLM asset management into an existing internal platform via OpenAPI
  • Running a Hugging Face-style workflow in an offline or air-gapped environment

Quick start

  1. 1

    Try the free SaaS version on the OpenCSG website

  2. 2

    Follow the official documentation to install with Docker Compose

Things to know

  • Self-hosting requires a Docker Compose or Kubernetes setup, which involves significant effort.
  • The README's quick start points mainly to a hosted SaaS service and external documentation rather than a single local command.

In collections

Self-hosted apps

Topics

#llm#model-management#huggingface#self-hosted#on-premise#dataset#mlops#open-source-platform#ai#asset-management

FAQ

What is csghub?
Open-source, self-hostable platform for managing LLMs, datasets, spaces and code, positioned as a private, on-premise alternative to Hugging Face. CSGHub is an open-source platform for managing large language model assets. It lets teams upload, download, store, verify and distribute LLMs such as DeepSeek and Llama, along with datasets, spaces and code, through a web interface, the git command line, a natural language chatbot, or the CSGHub SDK.
Is csghub free to use? What license does it have?
csghub is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
Is csghub still maintained?
Actively maintained. The last commit to csghub was on Nov 5, 2026.
What can I use csghub for?
Hosting an internal private model registry for an organization that cannot use public cloud hubs Managing LLM versions, datasets and deployed applications in one place Integrating LLM asset management into an existing internal platform via OpenAPI Running a Hugging Face-style workflow in an offline or air-gapped environment
How do I install csghub?
Try the free SaaS version on the OpenCSG website Follow the official documentation to install with Docker Compose
Where can I get csghub?
The source code and installation instructions are on GitHub: https://github.com/OpenCSGs/csghub

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects