kooka-server

Local inference web server for MLX / mlx-lm.

Note: mlx-lm==0.30.2 depends on transformers==5.0.0rc1, so uvx needs pre-releases enabled:

export UV_PRERELEASE=allow

Quickstart (single machine)

uvx kooka-server serve --model mlx-community/MiniMax-M2.1-4bit --host 127.0.0.1 --port 8080

See docs/DISTRIBUTED_SERVER.md.

Name		Name	Last commit message	Last commit date
Latest commit History 10 Commits
.github/workflows		.github/workflows
docs		docs
scripts		scripts
src/kooka_server		src/kooka_server
tests		tests
.gitignore		.gitignore
AGENTS.md		AGENTS.md
LICENSE		LICENSE
README.md		README.md
pyproject.toml		pyproject.toml