winged grotto
Hey folks
Quick question for anyone building ai-heavy apps with Next.js on Vercel.
How do you handle long-running API routes (like generating lots of content with OpenAI) when they exceed Vercel’s serverless timeout?
Should I run a separate Node server for heavy backend work? Or use something like Inngest/QStash? Or is there another common solution I’m missing?
Just trying to understand what the usual setup looks like. Thanks!