# Cursor usage data shows high coding output and a 10-to-1 input-to-output token ratio

DevFeed: [Cursor usage data shows high coding output and a 10-to-1 input-to-output token ratio](<https://devfeed.tech/articles/the-pulse-interesting-ai-coding-stats-from-cursor-40923.md>)

Original publisher: [Read original article](<https://blog.pragmaticengineer.com/the-pulse-interesting-ai-coding-stats-from-cursor/>)

Author: Gergely Orosz

Published: 2026-07-09T17:20:34Z

Content type: article

Language: en

Sources: [The Pragmatic Engineer](<https://devfeed.tech/sources/the-pragmatic-engineer-2.md>)

Topics: [cursor](<https://devfeed.tech/topics/cursor.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [ai-coding](<https://devfeed.tech/topics/ai-coding.md>), [data](<https://devfeed.tech/topics/data.md>), [Development](<https://devfeed.tech/topics/development.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [ai-coding](<https://devfeed.tech/tags/ai-coding.md>), [coding](<https://devfeed.tech/tags/coding.md>), [cost](<https://devfeed.tech/tags/cost.md>), [cursor](<https://devfeed.tech/tags/cursor.md>), [data](<https://devfeed.tech/tags/data.md>)

## AI overview

The article discusses Cursor's report based on two years of aggregated usage data. It describes large differences in code output between median, high-percentile, and top users, and reports that input tokens make up most usage because Cursor reads existing code and documentation.

## Source excerpt

Power users generate 10x as many lines of code vs the median, most of the AI spend is coming from input tokens not output ones, and almost half of AI changes are accepted without manual review by devs (!!)