Every speed-up is exact: the output is what plain decoding gives, token for token. Kinetic can also write a receipt for an answer, so anyone can rerun it and check it bit for bit. How it works, with…
Open source inference engine that tunes its kernels on your hardware to find the fastest way to run each model. 2x faster than Ollama. Works on Apple Silicon, NVIDIA, AMD, Intel, or just a CPU.
Local AI app and inference engine for agents. Run open-weight LLMs locally — private, 100% offline on your computer. Join our Discord: https://discord.com/invite/8wGSsvmg4V