Perplexity just shipped Computer Analytics, a new dashboard that lets you see exactly where your credits are going across every model Computer uses. If you have ever run a multi-step agent task and had no idea why your balance dropped by 200 credits, this is the feature you have been waiting for.
The problem it solves
Perplexity Computer breaks requests into tasks and subtasks, spins up sub-agents, and selects the best model for each step. That flexibility is powerful, but it creates a real visibility gap. Perplexity has built in real spending controls, but the difference between using Computer informed and using it blind is potentially hundreds of dollars a month.
In independent testing, simple tasks cost under 40 credits, research-heavy tasks cost 50 to 70 credits, and reusable automations cost 100 or more. With no per-model breakdown, users had to infer costs by watching the counter move. Computer Analytics replaces that guesswork with actual data.
What you get
The new Analytics section, found under Account Settings, surfaces credit consumption as a time-series broken down by model and credit source. Credit usage tiles show consumption this cycle split into Text, Image, Video, and Audio, with a per-thread log showing which queries used the most credits.
Don't miss what's next in AI
Join 300,000+ engineers and researchers who get the signal, not the noise.
- Full access to in-depth AI research breakdowns
- Be the first to know what's trending before it hits mainstream
- Daily curated papers, repos, and industry moves