Open-Weight Model
A model whose trained weights are published for download (e.g., Kimi, DeepSeek, Llama, Qwen), allowing self-hosting, fine-tuning, and inspection, as opposed to proprietary models accessible only through a vendor's API. Licenses vary, and openness shifts the operational and compliance burden to the deployer.
Want the full picture? This term comes from GenAI for Business, a complete free book on generative AI strategy and implementation by Prof. Shubin Yu (HEC Paris).