--- title: "Tokenmaxxing" description: "I burnt 250M tokens in a day. Tokenmaxxing is the deliberate practice of maximizing AI token consumption through parallelization." categories: ["AI","productivity"] keywords: ["tokenmaxxing","AI productivity","AI agents","parallel workflows","METR research","AI token consumption","agent orchestration"] ai_summary: "Tokenmaxxing is the deliberate practice of maximizing AI token consumption to boost productivity. By parallelizing agent workflows with a daily plan, the author processed 250 million tokens in a single day, a 20x increase in six weeks." date: 2026-04-01 lastmod: 2026-07-20 canonical_url: https://www.tomtunguz.com/tokenmaxxing/ author: "Tomasz Tunguz" --- Two days ago, I burnt 250 million tokens in a single day. That's up 20x in six weeks. This idea, called tokenmaxxing, is the deliberate practice of maximizing token consumption. The question : how much electricity can we turn into useful work? The secret is parallelization. Structure a plan at the start of the day that allows multiple agents to work simultaneously. [METR research](https://metr.org/blog/2025-03-19-measuring-ai-ability-to-complete-long-tasks/) shows the latest models can now work autonomously for 12 hours, up from 1 hour a year ago. Here's the ramp once I started implementing a [daily plan](https://tomtunguz.com/filling-the-queue-for-ai/) : ![The Path to 250M Tokens per Day](https://res.cloudinary.com/dzawgnnlr/image/upload/q_auto/f_auto/tokenmaxxing_ggplot.png) So, what did I do two days ago? Here's one example. I prepared a presentation for the [AI Engineers Tech Talk](https://luma.com/1xn4wqtm) on the infrastructure for building with agents that I'm delivering tonight. One agent pulled git commit history from the code repository & generated a lines-of-code chart. Another queried the agent error logs & built a time series of agent failures by root cause. A third fact-checked the METR research citations. A fourth built the presentation using a JavaScript library. A fifth critiqued the overall flow & content. All of this happened in the background. This was just one of the parallel flows in a day. The productivity ceiling? Still unmaxxed.