What is the difference between an LLM and a chatbot?

A chatbot is any program that converses with users; many older ones follow scripted rules and decision trees. An LLM is a large language model that generates responses statistically from training, so it handles open-ended, unanticipated input. An LLM can power a chatbot, but plenty of chatbots are not LLMs — they just match keywords and reply from a script.

How can you tell if you are talking to an LLM or a rule-based bot?

Rule-based bots break on anything outside their script: typos, rephrasing, or off-topic questions trigger fallback messages like 'I did not understand that.' LLMs handle novel phrasing gracefully, stay coherent across context, summarise, translate, and reason. Repetitive canned replies and rigid menus signal rules; fluid, adaptive answers signal an LLM.

Are rule-based chatbots obsolete?

No. Rule-based bots remain useful where answers must be exact and controlled — appointment booking, simple FAQs, regulated flows — because they never improvise or hallucinate. LLMs are better for open-ended help but need guardrails. Many production systems combine both, using rules for critical paths and an LLM for free-form questions.

Is this quiz suitable for beginners?

Yes. It is aimed at beginners and explains the reasoning behind every answer, so even if you guess wrong you learn the tell that distinguishes the two systems. By the end you will reliably recognise whether a given AI behaviour comes from a flexible language model or a scripted bot.

LLM vs Chatbot Quiz: Can You Tell the Difference?

Spot the difference

Not every chatbot is an LLM, and not every clever-sounding reply comes from a language model. Rule-based bots follow scripts and decision trees; LLMs generate responses statistically and handle input nobody anticipated. The difference shows up in how they behave — especially when you push past the happy path. This quiz drops you into ten real scenarios and asks you to call it: LLM or scripted chatbot.

How it works

Each question describes a short exchange or behaviour and offers a few choices. Pick the one you think is correct, then submit to see whether you were right along with a short explanation of the tell — the specific clue that gives the system away. At the end you get a score and a list of concepts worth revisiting. There is exactly one best answer per question.

Tips for spotting the tells

Watch how a system handles the unexpected: rule-based bots stumble on typos, rephrasing, and off-topic questions, falling back to canned “I did not understand” messages, while LLMs adapt smoothly. Notice coherence across a long conversation, the ability to summarise or translate on request, and willingness to reason about novel problems — all hallmarks of an LLM. Conversely, rigid menus, identical repeated replies, and a refusal to step outside a narrow domain point to a script. Keep these patterns in mind as you work through the questions.