# Test-Time Compute

> The work a model does while it answers. More of it can mean better answers.

**Field:** Artificial Intelligence  
**Category:** LLMs  
**Also known as:** Inference-Time Compute, Test-Time Scaling, Inference Scaling

## Think of it like

It is like taking more time on a hard test question. The more you check, the more you get right.

## Definition

Test-time compute is the work a model does while it answers, not while it is trained. A model can spend more of it by thinking longer, or by writing many answers and picking the best. On hard tasks, more of it often gives better answers, even with the same model. That is why labs now improve models in two ways. They train them bigger, and they let them think longer.

## Example

A team runs the same model twice on a hard maths set. With a larger thinking budget, it solves more of the problems, but each answer costs more.

## Why it matters

It is a second way to get better answers. You pay more per question, not to train a bigger model.

## Common mistake

Thinking more is always better. Past a point the gains flatten, and the cost keeps rising.

## Three principles

- **At answer time.** Not during training.
- **Think or sample more.** Longer steps, or many tries.
- **Gains then a plateau.** Cost keeps climbing.

## Related terms

- [Extended Thinking](https://learn.benibix.com/term/ai/extended-thinking)
- [Reasoning Model](https://learn.benibix.com/term/ai/reasoning-model)
- [Scaling Laws](https://learn.benibix.com/term/ai/scaling-laws)
- [Inference (AI)](https://learn.benibix.com/term/ai/inference)
- [LLM Sampling](https://learn.benibix.com/term/ai/sampling)

## Taught in

- [Reasoning Models and Thinking](https://learn.benibix.com/learn/ai-llms/reasoning-models): LLMs, 12/15

![Two rows. Thinking off: a question goes straight to an answer. Thinking on: the question goes to a highlighted box of thinking tokens, where the model plans, tries, checks and fixes, paid for like any other token, and then to the answer. A note says the extra tokens are the test-time compute, and the budget sets how many it may spend.](https://learn.benibix.com/diagrams/ai2-thinking.svg)

---

Source: https://learn.benibix.com/term/ai/test-time-compute
Author: Behnia Zarei (a developer in Vienna, Austria, who builds and teaches data, AI and business intelligence at BENIBIX)
Publisher: BENIBIX Learn (https://learn.benibix.com/)
Last updated: 2026-11-18
Licence: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Quote freely with attribution.
