Popular repositories Loading
-
project-maya
project-maya PublicProject 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, quan…
-
peasant-smith
peasant-smith PublicMaking new AI run on old hardware. Authoritative database for reproducible local-AI hardware benchmarks.
HTML 5
-
Strata
Strata PublicForked from Niko1221/Strata
Qwen3.8-Flash-Next on any consumer hardware: one-click install for Windows / Linux. Strata inference engine, OpenAI/Anthropic API on localhost, optional image input.
C++ 1
If the problem persists, check the GitHub status page or contact support.
