CodeBurn — Track AI coding token usage and cost across 37+ tools by model, project, and task.
Analyzed by Sai Pavan Gopularam · AI · Developer Tools · View on GitHub
- Stars: 11671
- Forks: 906
- Commits last 30 days: 100
- Health: Active (100 commits this month)
- Language: TypeScript
- License: MIT
What It Is
Imagine your monthly electricity bill, but instead of just a total, it shows you exactly how much each appliance used, which rooms consumed the most, and even where you left lights on unnecessarily. CodeBurn does this for your AI coding tools, reading session files locally to break down token usage and cost.
Most AI service bills are a black box, giving only a total. This tool solves the problem of opaque AI spend by providing granular insights, helping developers and teams understand where their AI budget goes, identify waste, and optimize their workflow to save money.
License Verdict
MIT License — Build and Sell Freely — Commercial Use Approved • No Copyleft Restrictions
The MIT license allows you to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the software. You can incorporate CodeBurn into proprietary products and services without needing to open-source your own code.
How to Use It
Get started instantly by running a single command to see your AI spend. For persistent use, install it globally via npm or Homebrew.
Prerequisites:
- Node.js 22.13+
Estimated setup time: 2 minutes.
npx codeburn
# To keep it:
npm install -g codeburn
# Or:
brew install codeburn
codeburnWhat I'd Build With This
AI Spend Audit & Optimization for Freelancers (micro-saas)
A web-based service where independent developers or small agencies upload their CodeBurn session data (or connect a local CodeBurn instance) for a detailed, interactive audit of their AI coding spend. It would generate custom reports highlighting cost-saving opportunities, model efficiency comparisons, and project-specific breakdowns. Who pays: Freelance developers, small dev shops, indie hackers. How to reach: Targeted ads on developer forums, Reddit communities, Product Hunt launches, content marketing around "saving money on AI tools."
Effort: 3 Weeks Build Time · Target: Indie Hackers · Pricing: $29/month
Team AI Cost Management Dashboard (saas)
A centralized SaaS platform for development teams to aggregate and visualize their collective AI coding spend. It would integrate with individual team members' CodeBurn instances, offer team-wide budget tracking, alert managers to overspending, identify top AI spenders/projects, and provide features for internal cost allocation. Who pays: CTOs, Engineering Managers, or FinOps teams in tech companies. How to reach: Direct sales, attending developer conferences, LinkedIn outreach to engineering leadership, partnerships with AI tool providers.
Effort: 3 Months Build Time · Target: Mid-sized Tech Companies · Pricing: $200-$1000/month
AI Governance & Chargeback System (enterprise)
A robust, on-premise or private-cloud solution for large enterprises to implement comprehensive AI cost governance. It would extend CodeBurn's capabilities to integrate with internal project management systems, HR systems for user identification, and financial systems for automated chargebacks to specific departments or cost centers. Includes advanced security, compliance reporting, and custom policy enforcement for AI model usage. Who pays: Enterprise IT departments, CIOs, Procurement. How to reach: Enterprise sales cycles, proof-of-concept engagements, consulting partnerships, direct outreach to Fortune 500 companies.
Effort: 6 Months Build Time · Target: Large Enterprises · Pricing: Custom annual contracts, starting at $50,000+
Sai Pavan Gopularam's Take
This is a smart tool for anyone serious about optimizing their AI coding spend. The local-first approach means privacy is baked in, which is a huge plus for developers. I could see myself paying $19/month for a simple web dashboard that syncs my local CodeBurn data and gives me aggregated insights without needing to keep the desktop app open.
Watch Out For
- Local Data Only: CodeBurn processes data locally on your machine. There's no built-in cloud sync or centralized dashboard, meaning you'd need to build infrastructure to aggregate data from multiple users.
- Requires Local AI Tool Usage: The tool only tracks AI coding activity that generates session files on your local disk. It won't track API calls made directly from server-side applications or cloud-based AI services without a local client.
- Node.js Dependency: While `npx` makes it easy to run, for persistent or advanced setups, a compatible Node.js environment is required, which might be an extra dependency for some users or systems.
I break down trending repos like CodeBurn every week — join the newsletter.