Agentic Rate Card

A simple server-rendered Next.js rate card for estimating agentic AI work across research, document creation, coding, design iteration, model stacks, GPU infrastructure, and multi-agent workflows.

The page is content-first and rendered on the server for search engines and a fast first paint, with no client-side data dependency.

Contents

• app/page.tsx — server-rendered rate card content
• app/globals.css — responsive editorial layout
• AI_Work_Rate_Card.html — legacy standalone one-page export
• AI_Project_Token_Cost_One_Page.md — Markdown companion
• AI_Project_Token_Cost_Guide.docx — longer-form document

The SSR app is also published at agentic-ratecard.vercel.app.

Run locally

npm install
npm run dev

The local development server runs at http://localhost:3002 so it does not collide with other projects using port 3000.

For a production check:

npm run build
npm run start

Refresh model pricing

The calculator reads its model rates from data/model-prices.json. To review the official provider pages before changing a rate, run:

npm run sync:prices

This creates a local data/pricing-source-snapshot.json with the latest price-related source snippets. Review it, update data/model-prices.json, and commit the revised rate file. The generated snapshot is intentionally ignored by Git so a provider-page layout change never silently changes published prices.