What is an AI model?
Updated · By Robert Breen
An AI model is the trained program that actually does the thinking inside an AI tool: it takes an input, like a prompt or a photo, and produces an output. Providers offer several models that differ in speed, cost and ability, and many tools let you pick one.
Why it matters for a small business
When you build with AI, picking the model is one of the few settings that changes cost and quality at the same time. A small, fast model is cheap and fine for short follow-up messages. A larger or more capable model costs more per request but may handle images, long documents or tricky instructions better. Paying for the biggest model on every simple job is a common way to waste money.
Models also change. Providers release new versions and retire old ones, so a workflow that names a specific model may need updating later. Knowing where the model is set in your tools makes that a two-minute change instead of a mystery.
In a real lesson: Build an AI Job Follow-Up Agent for Electricians
In the electrician follow-up agent lesson, you build an n8n agent for Oak & Wire Electric, a made-up electrical contractor, that writes follow-up messages after completed jobs. The model choice is its own step: under the AI Agent you click the + under Chat Model, add OpenAI Chat Model, connect your credential, then open the Model dropdown.
You pick gpt-5-mini. The voice-over gives the reasoning: it is fast, cheap and plenty for writing follow-up messages, and you should not reach for the expensive model until you have a reason. The job here is short text for three finished jobs (a 200-amp panel upgrade, a Level 2 EV charger and a whole-home surge protector), which a small model handles well.
The model is separate from everything else on the canvas. The system prompt, the Simple Memory and the Google Sheets Tool would all stay the same if you swapped in a different model later.
Try this lesson free or read the step-by-step guide.
Common confusions
AI model vs AI tool
ChatGPT, Claude and Gemini are tools (apps). Each runs on one or more models, and the model names change more often than the app names. In n8n you see the model directly, in the chat model node's Model dropdown.
Bigger model vs better result
A more capable model helps when the task is hard: images, long documents, many rules at once. For short, well-instructed text jobs, a clear prompt with your facts matters more than model size.
Tips
- Match the model to the job: the receipt extractor lesson switches to gpt-4o only because that agent has to read a photo.
- Test a cheaper model first with five real examples before paying for a larger one.
- Note which model each workflow uses so you can update it when the provider retires a version.
Related terms
Where you use it: free lessons
- Build an AI Job Follow-Up Agent for Electricians (n8n, 12 min)
- AI Receipt Extractor: Receipts to Google Sheets with n8n (n8n, 12 min)
- n8n AI Agent Tutorial: Save Social Media Ideas to Google Sheets (n8n, 12 min)
Frequently asked questions
- Can I change the AI model after my workflow is built?
- Yes. In n8n, open the chat model node and pick a different option in the Model dropdown. The rest of the workflow stays the same, but test it again because different models can word things differently.
- Does a more expensive model always write better?
- Not for simple jobs. Short emails and follow-ups usually come out just as well from a small model when the prompt includes your real details.