What does 'temperature' mean in AI settings and how does it change the output?

Temperature is a setting in many AI tools that controls how random or creative the model's responses are, with lower values producing more predictable, focused output and higher values producing more varied, sometimes surprising results. Think of it like a dial between caution and wildness. At temperature 0 or close to it, the model almost always picks the most likely next word. Ask it to complete 'The sky is...' and you'll get 'blue' every time. At temperature 1 or higher, the model is willing to pick less likely words, so you might get 'The sky is a bruised purple, heavy with the threat of rain.' Neither is wrong. They serve different purposes. Most AI tools default to a temperature around 0.7, which balances coherence with variety. Here's a concrete example: if you're using an AI tool to draft a legal disclaimer, set temperature low — around 0.2 — so the language stays precise and conservative. If you're brainstorming taglines for a coffee brand, crank it up to 0.9 or 1.0 and you'll get unexpected combinations like 'Coffee that argues back' alongside safer options. One thing beginners often miss: temperature doesn't make the model smarter or dumber. It just changes how adventurous the word choices are. A high temperature won't fix a vague prompt, and a low temperature won't save a poorly defined task. The setting works in combination with your instructions. If you find the output too bland, bump it up slightly rather than rewriting the prompt. If it's too scattered, dial it down. Some tools hide this setting behind an 'advanced options' menu, but it's worth finding. For more on how prompts and settings work together, see our guide on [how to write AI prompts](/blog/how-to-write-ai-prompts). **Related**: What's the difference between temperature and top-p? | How do I choose the right AI settings for blog writing?