Glossary

Foundation Model

A foundation model is trained on broad data and can be adapted to many tasks. Learn why a base model differs from the AI product that uses it.

Diploria

A foundation model is an AI model trained on broad data that can serve as a starting point for many tasks. Developers can adapt it through instructions, fine-tuning, or systems that supply tools and external information. Foundation models may process language, images, audio, or combinations of these.

A foundation model is not the same as a complete AI product. The product also determines the interface, search access, instructions, safety controls, and other behavior around the model.

Practical relevance

Two assistants built on the same foundation model can give different supplier recommendations if one searches current web pages and the other answers without retrieval. Their model may be shared, but their available evidence and instructions are not.

For visibility research, record the product and retrieval mode as well as model information when available. A model name alone does not fully describe the environment in which a brand appeared.

Learn more

Frequently asked questions

Know exactly where
AI mentions your brand

Track ChatGPT, Gemini, Perplexity, Claude and 8 more AI engines - weekly. See how your competitors rank, spot gaps, and fix them fast.