← Glossary

Glossary

Fine-tuning

Fine-tuning is further training of a pre-trained model on a focused dataset to adapt its behavior, style, or skill. Unlike RAG, it changes the model's weights rather than supplying knowledge at query time.


Fine-tuning is best for consistent format, tone, or a narrow task — not for keeping knowledge current, which RAG does better and more auditably. It carries a higher upfront cost and the risk of 'catastrophic forgetting' of general ability.

Most teams start with prompt engineering and RAG, and fine-tune only when prompting plateaus on a stable, well-defined task.

Related terms

Learn more