# Ask HN: When is fine-tuning a small LLM worth it?

> Source: <https://news.ycombinator.com/item?id=49807413>
> Published: 2026-09-22 20:11:30+00:00

| [Guidelines](newsguidelines.html) \|[FAQ](newsfaq.html) \|[Lists](lists) \|[API](https://github.com/HackerNews/API) \|[Security](security.html) \|[Legal](https://www.ycombinator.com/legal/) \|[Apply to YC](https://www.ycombinator.com/apply/) \|[Contact](mailto:hn@ycombinator.com) | 

|  | **[Hacker News](news)**[new](newest) \|[past](front) \|[comments](newcomments) \|[ask](ask) \|[show](show) \|[jobs](jobs) \|[submit](submit) | [login](login?goto=item%3Fid%3D49807413) | 

|  |  | [Ask HN: When is fine-tuning a small LLM worth it?](item?id=49807413) | 
|  |  | 4 points by [kooldeep7](user?id=kooldeep7)[13 minutes ago](item?id=49807413) \|[hide](hide?id=49807413&goto=item%3Fid%3D49807413) \|[past](https://hn.algolia.com/?query=Ask%20HN%3A%20When%20is%20fine-tuning%20a%20small%20LLM%20worth%20it%3F&type=story&dateRange=all&sort=byDate&storyText=false&prefix&page=0) \|[favorite](fave?id=49807413&auth=bd45abcdda65bf55fc398baa9fe65e92113349d0) \|[discuss](item?id=49807413) | 
|  |  | I'd be interested in hearing about your experiences. What kind of task did you use it for, what model did you train, and what were the results? Feel free to share examples of what you've tried. |