Hacker News new | past | comments | ask | show | jobs | submit login
[flagged]
ent101 3 months ago | hide | past | favorite



Does anyone know if achieving this requires a deeply fine-tuned or modified model compared to 4o? At first glance, it feels like a small iterative change that would be achievable with prompt engineering and some slight code modifications alone.


Query limits: 30/week (50 for mini model). I wonder what the API pricing is.

(not a typo: that's PER WEEK)


It must be iteratively calling itself to check/re-check work. EXPENSIVE.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: