Install

Install the runtime. Apache-2.0.

The Kortecx runtime is open-source and installs today — one FFI-free binary, single-node by default. Pick an install method, then run your first agent in seconds.

kortecx :: v0.1.0

Install

curl -fsSL https://raw.githubusercontent.com/Kortecx/kortecx/main/scripts/install.sh | sh

Linux x86_64/arm64 · macOS arm64 · SHA-256 verified · no sudo

Verify

kx --version

The single kx binary is the runtime's entrypoint: run · replay · serve · digest and more.

Run the 60-second proof

Run a workflow, crash it mid-commit, then replay — and watch the digest come back identical. Committed work is re-read, not re-run: exactly-once delivery with transparent recovery.

Open the quickstart

Local inference

Local LLM inference is opt-in (Tier 1): it builds llama.cpp, so it needs a C++ toolchain. GPU offload uses Metal on Apple silicon; CUDA is cloud-only. The default install is FFI-free and needs none of this — reach for local inference only when you want a model running on your own machine.

Read the docs, or grab the source.

Read the docs GitHub

Need managed capacity? Join the cloud waitlist.