tech
Peter Steinberger's 100 AI agents racked up $1.3 million in OpenAI tokens in 30 days building OpenClaw
OpenClaw creator Peter Steinberger spent $1.3 million in OpenAI API tokens in 30 days running 100 Codex instances on his open-source project. The bill, covered by OpenAI where Steinberger now works, represents 603 billion tokens across 7.6 million requests and provides the most concrete public data point on the cost of autonomous AI coding at scale.
TL;DR
- Peter Steinberger's OpenClaw project generated a $1.3 million API bill in 30 days from running 100 Codex instances.
- The costs covered 603 billion tokens and 7.6 million requests, primarily using GPT-5.5.
- OpenAI, where Steinberger now works, is covering the bill as a research investment into AI development costs.
- Autonomous AI agents perform tasks like code review, security scanning, issue management, and feature development.
- The 'Fast Mode' pricing significantly increased costs; disabling it would reduce the bill to approximately $300,000 per month.
- Steinberger's previous success includes founding PSPDFKit, a widely adopted PDF rendering framework.
- OpenClaw became the fastest-growing open-source project on GitHub, aiming to change the world rather than build a large company.
- The high costs highlight the tension between subscription models designed for human interaction and the massive compute demands of autonomous agents.
- This bill provides a realistic baseline for enterprise teams evaluating the cost of deploying AI coding tools at scale.
- The rise of AI coding agents signals a shift in software development, with questions centering on cost and payment rather than AI's capability to write code.