Inkling is an open-weight multimodal Mixture-of-Experts model from Thinking Machines Lab, with 975B total parameters and 41B active parameters. It natively accepts text, image, and audio inputs and generates text outputs, supporting a context window of up to 1 million tokens. Inkling is designed for general-purpose reasoning, coding, agentic and tool-use systems, instruction following, and multimodal tasks, and is intended as a broad foundation model suitable for fine-tuning and domain adaptation. It was pretrained on 45 trillion tokens of text, images, audio, and video, and features a hybrid local/global attention architecture with controllable thinking effort.