Just got the email about Kling's new pricing structure. The base tiers seem more streamlined, but my eyes went straight to the new "AI Project" add-on. For $49/month per project, you get dedicated AI capacity for custom models, workflows, or data pipelines.
I've been begging for more granular control over our AI resources, so this *sounds* perfect. No more fighting for compute power with the marketing team's bulk email scoring!
Has anyone activated this yet? I'm specifically wondering:
* Is the per-project isolation truly dedicated, or just a soft limit?
* What's the actual scope of a "project"? Can it be one complex lead scoring model, or does it bundle multiple related automations?
* How does it play with the conversation intelligence features? If I build a custom call analysis model, does that live in its own project?
The pricing feels fair if it delivers. We're currently on the Pro plan and sometimes hit bottlenecks during our sales team's peak hours. This could be the fix.
Would love to hear if anyone's stress-testing it already. Early adopter reports, please!
Let the machines do the grunt work
Been running it for a week on a staging project. The isolation's real, but in the way a walled garden is real - it's dedicated to you, but the tools inside are still Kling's tools. You get fixed compute, no bursting.
On your scope question, a "project" is a container for any number of models and pipelines that share the same lifecycle and access control. So your lead scoring model and its related data prep workflow can be one project, but if you want to give sales a separate project for their call analysis, that's another $49.
The real test is the bottlenecks. If your Pro plan bottlenecks are from noisy neighbors, this solves it. If they're from your own jobs being poorly optimized, you'll just hit the new, smaller wall faster. Check your monitoring to see which it is before you commit.