Noob question: does Langchain run models locally within my app, or via third-party API?

[mitya33]

I'm new to Langchain. Could someone please confirm whether completions are being generated locally within my app, i.e. by the package, or via third-party API? I presume the latter, hence the need for an API key relating to the model you're using, however when I asked the GPT-powered Langchain assistant, it said the former.

Additionally, is there any way I can use Langchain with Cloudflare AI, which hosts and runs models on its side?

Thank you in advance.

[mitya33]

Figured this out myself.