Technology
LLM (Large Language Model)
The transformer-based language model under every AI scribe.
Every modern ambient scribe is built on a large language model — GPT-, Claude-, Gemini-, Llama-class — for the note-generation step. Vendors differ in which foundation model(s) they use, whether they fine-tune, and whether they keep customer data out of training. The latter (training-on-data policy) is the more consequential question for buyers.
See also
- Training on customer data — Whether the vendor uses your sessions to train its AI models.
- AI hallucination — When an AI model writes something confidently into a note that was never said.