san-digital/tokspur: Run local Qwen3.8 Flash-Next inference on macOS

Tokspur is an open-source application that runs the Qwen3.8 Flash-Next model locally on macOS using a native Metal worker. It provides an OpenAI-compatible chat API, requiring an Apple M5 Max with 128 GB of unified memory. The software is distributed under the Apache 2.0 license. Open article →