Agnes AI Usage Analytics
Upload one Agnes usage CSV to analyze text tokens, image counts, and video seconds side by side, with requests and cost as supporting context
Drop your Agnes usage CSV here or click to upload
Current version supports a single Agnes usage CSV file
Files stay in your browser — nothing is uploaded
How It Works
1. Export Data
Go to Agnes AI usage export and download the usage CSV for the period you want to analyze.
2. Drag & Drop
Drag the Agnes usage CSV onto this page, or click the upload area to select the file manually.
3. View Analytics
Instantly inspect text tokens, image counts, and video seconds side by side, then compare requests, cost, per-key breakdowns, custom project groupings, and trends — all processed locally in your browser.
Frequently Asked Questions
No. All CSV parsing and cost computation runs entirely in your browser. Your data never leaves your device.
Upload a single Agnes usage CSV file. The current version does not require ZIP extraction or multi-file pairing.
Not yet. The current Agnes version focuses on one CSV at a time so the workflow stays simple and reliable.
Any model listed in your Agnes usage export is supported. The dashboard always surfaces text tokens, image counts, and video seconds side by side for the selected model or for all models combined.
That usually means the current CSV, model filter, or date range has no records for that modality. Zero text tokens, zero images, or zero video seconds are all valid outcomes.
In the Agnes version, this usually means the uploaded file is empty, missing required columns, or is not a valid Agnes usage CSV.
Check the Troubleshooting section in our full User Guide. It covers common issues like CSV format errors, missing columns, and upload validation.
Yes, each file must be under 50 MB. This keeps the browser responsive and prevents oversized files from freezing the page.
Yes. Switch to the 'By Project' tab and click the gear icon to open the configuration panel. Drag API key names from the unassigned pool into your custom project groups. Your project configuration is saved in your browser's local storage.
About
Why We Built This
As our Agnes AI usage kept growing, the raw CSV became harder to interpret at a glance, especially once text, image, and video records started appearing in the same export. To answer questions like "which project spent the most?", "which key produced the most images?", and "where did video generation spike?", we built a browser-side dashboard that keeps all three core dimensions visible together.
Under the Hood: Privacy & Tech
When handling billing and usage data, privacy is non-negotiable. That's why we chose a pure frontend architecture. Built with Next.js 16 (App Router) and React 19, powered by Papa Parse and ECharts, the app parses one Agnes usage CSV locally in your browser, counts only success rows, and keeps four focused dashboard tabs for overview, projects, keys, and trends. No backend server, no database — your data never leaves your device.
About MindRose
This project is open-sourced by Gavin Chen and the MindRose team. MindRose is a tech team focused on delivering lightweight digital solutions for small and medium manufacturers, logistics companies, and cross-border traders. We don't sell vague "digital transformation" concepts — we use AI and full-stack agile development (Next.js, React, Mendix, etc.) to deliver applications that solve real business pain points within weeks. From SEO-optimized independent e-commerce sites, to agile internal system builds, to AI agent workflow integration — we understand tech, and more importantly, we understand your business.
Let's Work Together
Looking to build a custom data dashboard for your business? Need deep AI integration into your workflows? Want modern, elegant UI/UX? Need cross-department, multi-terminal business systems? Or interested in collaborating on this open-source project?
We offer professional bottleneck diagnosis, transformation analysis, tailored business solutions, and AI-powered development and operations.