2-5x faster fine-tuning, reinforcement learning and inference for LLMs, with an optional Studio web UI.
Local, OpenAI-compatible API server for running LLMs, image, audio and embedding models on your own machine - no GPU or cloud account required.
