Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
An Open-source, self-hosted AI model hub with Hugging Face compatibility, accelerating vLLM/SGLang performance.
| Date | Stars |
|---|---|
| 2026-07-24 | 257 |
| 2026-07-25 | 258 |
| 2026-07-28 | 258 |
| 2026-07-30 | 258 |
| 2026-08-06 | 258 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
35.0
growth rate 0.00%/day
# MatrixHub [](https://deepwiki.com/matrixhub-ai/matrixhub) [](https://www.bestpractices.dev/projects/13641) **MatrixHub** is an open-source, self-hosted AI model registry engineered for large-scale enterprise inference. It serves as a drop-in private replacement for Hugging Face, purpose-built to accelerate **vLLM** and **SGLang** workloads. ## 💡 Why MatrixHub? MatrixHub streamlines the transition from public model hubs to production-grade infrastructure: * **Zero-Wait Distribution**: Eliminate bandwidth bottlenecks with a **"Pull-once, serve-all"** cache, enabling 10Gbps+ speeds across 100+ GPU nodes simultaneously. * **Air-Gapped Delivery**: Securely ferry models into isolated networks while maintaining a native `HF_ENDPOINT` experience for researchers—no internet required. * **Private AI model Registry**: Centralize fine-tuned weights with **Tag locking** and CI/CD integration to guarantee absolute consistency from development to production. * **Global Multi-Region Sync**: Automate asynchronous, resumable replication between data centers for high availability and low-latency local access. ## 🛠️ Core Features ### 🚀 High-Performance Distribution * **Transparent HF Proxy**: Switch to private hosting with zero code changes by simply redirecting your endpoint. * **On-Demand Caching**: Automatically localizes public models upon the first request to slash redundant traffic. * **Inference Native**: Native support for **P2P distribution**, OCI artifacts, and **NetLoader** for direct-to-GPU weight streaming. ### 🛡️ Enterprise Governance & Security * **RBAC & Multi-Tenancy**: Project-based isolation with granular permissions and seamless LDAP/SSO integration. * **Audit & Compliance**: Full traceability with comprehensive logs for every upload, download, and configuration change. * **Integrity Protection**: Built-in malware scanning and content signing to ensure models remain untampered. ### 🌍 Scalable Infrastructure * **Flexible Storage**: Supports PVC-backed local file system and NFS storage today, with S3-compatible object storage planned for a future release. * **Reliable Replication**: Policy-driven, chunked transfers ensure data consistency even over unstable global networks. * **Cloud-Native Design**: Optimized for Kubernetes with official **Helm charts** and horizontal scaling capabilities. ## 🌐 Live Demo Try MatrixHub instantly at **[demo.matrixhub.ai](https://demo.matrixhub.ai/)** — no setup required. Sign in with the public demo credentials: | Username | Password | | --- | --- | | `admin` | `changeme` | > The demo is for evaluation only and may be reset at any time. ## 🚀 Quick Start ### Docker Compose Deployment Download the Docker Compose files for a released version and start MatrixHub: ```bash export MATRIXHUB_VERSION=v0.1.1 mkdir -p matrixhub && cd matrixhub curl -fL \ "https://raw.githubusercontent.com/matrixhub-ai/matrixhub/$MATRIXHUB_VERSION/deploy/docker-compose.yml" \ -o docker-compose.yml curl -fL \ "https://raw.githubusercontent.com/matrixhub-ai/matrixhub/$MATRIXHUB_VERSION/deploy/config.yaml" \ -o config.yaml MATRIXHUB_IMAGE_TAG="$MATRIXHUB_VERSION" docker compose up -d ``` For a newer stable release, replace `v0.1.1` with the version you want to run. If port `3001` is already in use, set `MATRIXHUB_HTTP_PORT` before starting the stack, for example `MATRIXHUB_HTTP_PORT=3002`. Open the MatrixHub web console: ```text http://127.0.0.1:3001 ``` Sign in with the default local credentials: | Username | Password | | --- | --- | | `admin` | `changeme` | Change the default password before exposing the instance outside your local machine. To stop the local stack: ```bash docker compose down ``` ### Helm (Kubernetes) Deployment #### Prerequisites Currently, the Helm chart supports PVC-backed storage for MatrixHub data. S3-compatible object storage is planned
Excerpt of 6,410 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:a777b3bc2230f3dd, topic:llm-inference, topic:vllm
matched fp:a777b3bc2230f3dd, topic:model-registry, readme:model registry
matched fp:a777b3bc2230f3dd, topic:kubernetes
matched fp:a777b3bc2230f3dd, topic:llm