gpustack

GPUStack is a specialized GPU cluster manager for orchestrating inference engines like vLLM and SGLang, directly relevant for deploying large models such as DeepSeek on multi-GPU clusters.

Open repository
Stars
5.1k
Updated
2026-06-07

Summary

GPUStack is a specialized GPU cluster manager for orchestrating inference engines like vLLM and SGLang, directly relevant for deploying large models such as DeepSeek on multi-GPU clusters.

GitHub repository 'gpustack/gpustack' with 5,111 stars and recent activity.. Description explicitly states it is a 'GPU cluster manager' for inference engines like 'vLLM and SGLang'.. Discovered via a 'deepseek-model-deployment' query, indicating direct relevance to deploying large language models.

Tags

Also appears in