Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
natoucs
searching Neon…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
5 ms
·
1.
▲
by
natoucs
2mo ago
tokentoken.club has it 30% cheaper at 2$ input
2.
▲
by
natoucs
10mo ago
Few reasons: keep access to the frontend features of each providers, have access to my chats I have in the individual frontend apps, to not have to trust a 3rd party provider, to not have to update the app each time a new model comes out
3.
▲
by
natoucs
10mo ago
It doesn't offer to keep access to the native frontends
4.
▲
by
natoucs
10mo ago
That makes sense. But you can't access the native frontend if it is in a webapp
5.
▲
by
natoucs
10mo ago
good idea! And yes that's the issue
6.
▲
by
natoucs
10mo ago
I would if I found a way to keep access to the frontends of each LLM provider while being the web
7.
▲
by
natoucs
10mo ago
Very nice! Do you still get access to the frontend of the original LLM providers and do you have to insert API keys ?
8.
▲
by
natoucs
10mo ago
iframe - you got it - it embeds the web apps
9.
▲
by
natoucs
7y ago
I think some people here are confused because they imagine financial/customer synthetic data where the pattern to simulate is unclear, instead of computer vision, where the pattern to replicate is obvious as we see it before our eyes.