Nerf – Definition & Meaning

Definition of nerf

: to reduce, or appear to reduce, the capability, reasoning effort, reliability, or usefulness of an AI model after it has been released
- Users claimed the coding model had been nerfed after it began stopping halfway through complex tasks.
- The benchmark showed no change in the weights, suggesting that the apparent nerf came from new defaults in the surrounding software.

Two meanings, one word

In everyday English

(in video games) to reduce the power or effectiveness of a character, weapon, or ability, usually through an update.

In AI

To reduce, or appear to reduce, the capability, reasoning effort, reliability, or usefulness of an AI model after it has been released.

Origin & history

first known use1995

The verb meant 'to make less effective' by 1995 and became common gaming slang for weakening something in a balance update. AI users later borrowed it for perceived post-release drops in model quality, especially when a service changed routing, reasoning settings, limits, safety behavior, or agent scaffolding.

Test yourself

Which of these is the meaning of nerf?

Related terms

quantization shrinking a model by storing its numbers with less precision (for example 4 bits instead of… model the trained program that results from machine learning: a set of learned numbers plus the… inference the act of running a trained model to produce an output, such as answering a prompt or… post-training everything done to a model after pre-training to make it useful and safe, such as instruction… harness engineering the work of building everything around an AI model that lets it act as a reliable agent… latency the delay between sending a request to a model and getting a response, often measured as time…

Cite this entry

"Nerf." AI Dictionary, Dadgogo, https://dadgogo.com/dictionary/nerf/. Accessed 4 Oct. 2026.