ModelCascade – Route LLM calls to your own GPU first, cloud second(github.com)
1 pointby wayneIA5 hours ago2 comments