# CSGHub

CSGHub is a brand-new open-source platform for managing LLMs, developed by the OpenCSG team. It offers both open-source and on-premise/SaaS solutions, with features comparable to Hugging Face. Gain full control over the lifecycle of LLMs, datasets, and agents, with Python SDK compatibility with Hugging Face. Join us! ⭐️

Repository: https://github.com/OpenCSGs/csghub
Canonical: https://ross.abutalabs.com/products/csghub
Homepage: https://opencsg.com
Language: Vue
License: Apache-2.0
License Family: permissive
Topics: ai, huggingface, llm, management-system, platform, asset-management, dataset, deepseek, deploy, finetune, git, inference, model, prompt, ray, space
Last push: 2026-08-12T06:49:27+00:00
Link (homepage): https://opencsg.com

## Health v2 (maintenance only)
Score: 91/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 97, release rhythm 96, longevity 68
- inputs: {"age_days": 964, "days_push": 21, "days_rel": 24, "gap_med": 29, "n_releases_24m": 26}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4097, forks 515 (observed 2026-08-28T04:08:35.394401+00:00)

## What it is
CSGHub is an open-source, self-hostable platform for managing the full lifecycle of LLM assets including models, datasets, spaces, and agents, comparable to a private on-premise Hugging Face. It provides a web interface, git CLI, chatbot, and Python SDK with Hugging Face compatibility, plus OpenAPIs and microservice submodules for enterprise integration.

## Use cases
- self-host a private hugging face alternative
- manage and version llm models and datasets on-premise
- deploy llm applications and model inference endpoints
- download and upload models via git or python sdk
- run an offline air-gapped model registry for enterprise
- sync models from hugging face to an internal platform
- fine-tune and serve open-source models like deepseek or llama

## When to choose
- you need full control over model/dataset assets behind your firewall
- you want Hugging Face-compatible workflows (SDK, git) on your own infrastructure
- you need enterprise access control, on-prem deployment, and OpenAPI integration

## When to avoid
- you just need a public model hub and are fine with huggingface.co
- you want a lightweight model server without asset management features
- you cannot operate a multi-service platform with docker/orchestration

## Facets
- artifact type: application
- maturity: active
- function: machine-learning, llm-inference, rag, agent-framework, web-framework, api-framework, self-hosted, developer-tools, file-upload, search-engine
- domain: large-language-models, machine-learning, artificial-intelligence, self-hosted, developer-tools, data-science, apis
- platform: self-hosted, cloud, python
- tags: model-registry, huggingface-alternative, llm-asset-management, model-hosting, dataset-management, on-premise-ai, openapi, git-based, model-deployment, spaces, docker, web-server, vue

## Member repositories
- OpenCSGs/csghub (main) score 91
- OpenCSGs/csghub-server (backend) score 92

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:35.394401+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:23:18.157144+00:00, confidence not recorded.
  - readme: https://github.com/OpenCSGs/csghub (fetched 2026-08-28T04:08:35.394401+00:00, sha ba4c509e92f3)
  - homepage: https://opencsg.com (fetched 2026-08-29T09:15:15.427518+00:00, sha f827daad2209)
- Data as of 2026-08-30T08:39:29.467469+00:00.