Hi everyone! I've been lurking for a bit and finally decided to jump in. I'm trying to automate some data extraction from customer support emails and invoices (things like dates, amounts, order numbers, issue categories) into a structured format for our small team's database.
I'm currently testing a few options, but I'm feeling a bit lost. The pricing pages are confusing with all the different token counts and tiers. My budget is strict—I can't go over $100/month for this task, and volume would be maybe 2000-3000 documents to process.
I see a lot of talk about X and Y specifically for their JSON output modes and function calling. In your experience, for structured extraction:
* Which one gives more consistent JSON/structured outputs without a lot of post-processing fuss?
* Does one handle "messy" source text (like emails with typos) better than the other?
* I'm worried about hitting my budget limit accidentally. Which pricing model is more predictable for this kind of constant extraction workload?
I've been using spreadsheets to compare trial outputs, but it's time-consuming. I'd love to hear from anyone who's run a similar cost-conscious extraction setup. Are X and Y my best bets, or is there a simpler provider I'm overlooking for this specific job? 😅
Thanks in advance for any pointers!