ModelRefs / Base Model — AI Glossary
Base Model — AI Glossary
A pretrained LLM that has not been instruction-tuned or aligned — it predicts the next token without following instructions.
Overview
Base models (e.g. Llama base, Mistral base) are the raw output of pretraining. They must be further fine-tuned (SFT + RLHF or DPO) to become useful assistants. Most practitioners start from a base model when building specialized systems.
Reference details
| Topic | models |
|---|---|
| Also known as | pretrained model, raw model |
| Last reviewed | 2026-06-24 |
Related terms
Commonly confused with
Base means pretrained and nothing more — no instruction tuning, no alignment. This produces a specific and much-repeated surprise: prompt a base model with a question and it may continue the text with more questions, because it was never trained to treat a prompt as a request. Instruction-tuned and chat variants of the same weights behave completely differently and are usually what people actually want.
Continue your research
Use these connected ModelRefs sections to compare alternatives, inspect implementation paths, and review the evidence and governance boundaries relevant to Base Model — AI Glossary.
Frequently asked questions
What is Base Model?
A pretrained LLM that has not been instruction-tuned or aligned — it predicts the next token without following instructions.
Is Base Model the same as pretrained model?
Yes — pretrained model, raw model are common aliases for Base Model.
What concepts are related to Base Model?
Closely related concepts include llm, pretraining, instruct model, sft.