# Caveman

59 Hacker News commenters recommended Caveman and 33 criticized it, November 2025 to November 2026. It was recommended in 7 of 13 months.

**Verdict:** People recommend it for shorter, plainer Claude replies that are quicker to read and less padded. Critics say the token savings are small and it may hurt reasoning quality.

## Why people recommend it

- Shorter, quicker-to-read LLM answers (7 comments): "it will make my day not to have to read through entire essays about some trivial solution" ([source](https://news.ycombinator.com/item?id=47647970)) "in the time I've actually come to like its matter-of-fact and short way of responding" ([source](https://news.ycombinator.com/item?id=49666629)) "anecdotally I do think it saves me quite a lot of time on reading LLM outputs" ([source](https://news.ycombinator.com/item?id=49659031))
- Terse, matter-of-fact writing style (5 comments): "in the time I've actually come to like its matter-of-fact and short way of responding" ([source](https://news.ycombinator.com/item?id=49666629)) "I already enjoy reading its output more than vanilla so suits me well." ([source](https://news.ycombinator.com/item?id=47793771)) "I've found it effective so far. Chat output isn't nearly as flowery." ([source](https://news.ycombinator.com/item?id=49473616))
- Cuts output tokens without losing accuracy (4 comments): "tools like RTK, Headroom, caveman mode and others do reduce input and output tokens that need to be processed" ([source](https://news.ycombinator.com/item?id=48592020)) "caveman is looking to be the only tool that actually doesn't regress on reasoning while taking costs down" ([source](https://news.ycombinator.com/item?id=49658835)) "It also saves you tokens and it has been linked with more accuracy." ([source](https://news.ycombinator.com/item?id=48905846))
- Fun and enjoyable to use (4 comments): "It really releases the stress slightly to call bugs, buggas and generally role play a humorous setting" ([source](https://news.ycombinator.com/item?id=47959286)) "Actually tried caveman mode yesterday and it made everything SO MUCH BETTER" ([source](https://news.ycombinator.com/item?id=47860090)) "I already enjoy reading its output more than vanilla so suits me well." ([source](https://news.ycombinator.com/item?id=47793771))

## What people warn about

- Token savings are small or negligible (6 comments): "I tested both for weeks and they didn't really saved that much money for Opus model." ([source](https://news.ycombinator.com/item?id=48082313)) "the benchmark suggests it doesn't offer as much benefit as" ([source](https://news.ycombinator.com/item?id=47956359)) "it does not reduce the token usage really" ([source](https://news.ycombinator.com/item?id=47648092))
- May hurt reasoning or accuracy (4 comments): "caveman is looking to be the only tool that actually doesn't regress on reasoning while taking costs down" ([source](https://news.ycombinator.com/item?id=49658835)) "In real sessions it doesn’t really save much longterm and made sessions more forgetful/take longer." ([source](https://news.ycombinator.com/item?id=49281495)) "changing something like this absolutely changes the behaviour of the models responses" ([source](https://news.ycombinator.com/item?id=47797199))
- Caveman speech is grating (3 comments): "it takes things too far in that direction, with "cave man" speech that sucks" ([source](https://news.ycombinator.com/item?id=49526002)) "caveman sounds bad and is not as easy to understand compared to normal English" ([source](https://news.ycombinator.com/item?id=47807693)) "Now I can’t stand caveman anymore. It is really frustrating working with it." ([source](https://news.ycombinator.com/item?id=49051882))
- Same as simply asking for brevity (3 comments): "for THAT value of "better" it's not much different" ([source](https://news.ycombinator.com/item?id=47863051)) "turns out I never read the words that caveman omits, so to me it makes zero difference" ([source](https://news.ycombinator.com/item?id=47956264)) "You can just add "be brief" to the prompt to replace the entire plugin. Same results." ([source](https://news.ycombinator.com/item?id=47993652))

## Alternatives hackers recommend

- [Compound Engineering plugin by Every](https://hackerapproved.com/p/compound-engineering-plugin.md): 9 recommend
- [pi-subagents](https://hackerapproved.com/p/pi-subagents.md): 8 recommend
- [Claudish-to-English plugin](https://hackerapproved.com/p/claudish-to-english.md): 7 recommend

## Cite as

Hacker Approved, "Caveman", https://hackerapproved.com/p/caveman/ (counts from 1,795,640 Hacker News comments, November 2025 to November 2026).