Hacker News Reader

Yea I am testing through OpenRouter - have you noticed 3.7 flash being significantly faster?

I guess it might be relative, but switching from VertexAI endpoint to OpenRouter was like 2-3x faster for us.