Project Maya - run GLM-5.3-Flash (321B MoE) privately on your own GPU(s), from one card to 16, NVIDIA or AMD, Linux or Windows. Engine, server, dashboard, OpenAI- and Anthropic-compatible API, quant tools.
↑ +27 today
Discover the best Glm 5 open source repositories on GitHub. Explore trending tools, live star counts, source code, and developer projects updated daily.