LLM vs. SLM (Large vs. Small Language Models)
NewBoth are language models; the difference is size, and with it cost, speed and where they can run. Large models are the most capable and usually run in a provider's data centre. Small models are far cheaper and faster, can run on a laptop, phone or a single server, and are often fine-tuned for one narrow job. Choosing a small model on purpose — and proving it was good enough — is a sign of cost-conscious AI engineering.
How to tell them apart on a resume
LLM (large)
GPT, Claude, Gemini, large Llama models, complex reasoning, general-purpose assistants, API costs measured per million tokens.
SLM (small)
Phi, Gemma, small Llama or Qwen models, on-device, edge, quantization, distillation, fine-tuned for one task, latency and cost savings.
The question that settles it
“Why did you pick the model size you used — and did you test whether a smaller, cheaper model would have done the job?”
Read the full definitions
Open the full tool for the other look-alike pairs, role profiles, and the JD decoder.