Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
dovin
9 months ago
|
parent
|
context
|
favorite
| on:
Benchmarking GPT-5 on 400 real-world code reviews
I don't consider myself a font snob but that web page was actually hard for me to read. Anyway, it's definitely capable according to my long-horizon text-based escape room benchmark. I don't know if it's significantly better than o3 yet though.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: