Hey everyone! I've been using OpenPipe for a few months now to handle some of our lighter ETL workloads and prototyping, and I was really loving it. The model routing and cost optimization features were a game-changer for my team's self-serve analytics projects.
I just got the email about their new pricing structure, and I have to say, my jaw dropped a little 😅. The move away from the simpler, usage-based tiers to this new model with dedicated throughput (CU) commitments feels like a huge jump. For a small team like ours, it seems like we'd be locking into a minimum spend that's way above what we currently use.
Is anyone else feeling this shift? I'm trying to understand:
* **For those on the new plan:** How are you estimating your needed throughput? I'm worried about over-provisioning.
* **Are there hidden costs** we should watch out for, compared to the old pay-as-you-go?
* **For similar use cases (light data pipelines, dbt model integrations),** are you sticking with OpenPipe or starting to look at alternatives?
I really want to keep using it because the product itself is fantastic, but the pricing now feels like it's geared more towards larger enterprises. As someone still getting their feet wet with LLM pipelines, I'd love some perspective or maybe recommendations for other tools that might fit a beginner-to-intermediate stage better. Thanks in advance!