Skip to content
Notifications
Clear all

Kling vs. building a simple RAG system ourselves. Cost analysis for year one.

1 Posts
1 Users
0 Reactions
33 Views
(@j_carter)
Estimable Member
Joined: 6 months ago
Posts: 113
Topic starter   [#20990]

Hi everyone — I've been lurking for a bit, but this is my first post. I'm currently evaluating Kling for our small team, and the big question we're stuck on is whether to go with a managed solution like Kling or build a simple RAG system in-house.

We’re a team of 10, primarily using Google Workspace and a lightweight CRM. Our use case is straightforward: we want to query our internal documentation (Google Docs, some PDFs, and past support tickets) to get quick answers. We don’t need heavy customization, but reliability and reasonable cost are key.

I’ve done a rough cost analysis for the first year, and I’d love to get your thoughts on whether my estimates are in the ballpark.

**Option 1: Using Kling**
* Based on their public pricing, the Pro plan seems sufficient for our volume.
* Estimated annual cost: ~$3,000.
* This includes setup, maintenance, and support. The main value is the time saved on development and operations.

**Option 2: Building a simple RAG ourselves**
* Development time: 2-3 weeks of a developer’s time (our biggest cost).
* Cloud infrastructure (embedding model API, vector DB, compute): ~$100-$150/month.
* Ongoing maintenance & tweaking: ~5 hours per month.
* Estimated first-year total: ~$15,000 - $20,000 when factoring in the initial development salary burden.

For us, the math heavily favors Kling, unless we’re missing a major hidden cost or a compelling reason to own the stack. Has anyone else run similar numbers? Did you encounter unexpected costs or challenges with either approach that changed your calculation?


Migration is never smooth.


   
Quote