Uber burned through its entire 2026 AI budget by April. The company rolled out Anthropic's Claude Code to roughly 5,000 engineers in December, watched usage double by February, and ran out of money before the first quarter ended.
Per-engineer token costs hit $500 to $2,000 per month. Ninety-five percent of Uber's engineers were using AI coding tools monthly. Eleven percent of the company's live backend code was being written by AI agents. When Uber's COO Andrew Macdonald was asked about the return on that spending, his answer was blunt: "That link is not there yet."
A $140 billion technology company with sophisticated financial infrastructure could not draw a line between its AI spend and a single consumer feature shipped to customers. That is the sentence worth reading twice.
Uber is not alone. Ramp's April 2026 AI Index reported that monthly AI token spend across enterprise customers grew 1,001 percent from January 2025 to April 2026. The median company now dedicates nearly 15 percent of its software budget to AI tools. A Priceline engineer burned $40,000 in tokens in a single month. EY estimates that a standard chatbot interaction costs roughly $0.04 — but an orchestrated agentic workflow, where AI models call tools, spawn sub-agents, and iterate across multiple reasoning steps, costs approximately $1.20 per interaction. That is a 30x multiplier, and it is built into the architecture of where enterprise AI is heading.
Uber has since imposed a $1,500 monthly spending cap per employee. Meta reportedly dealt with "tokenmaxxing" — employees intentionally overusing AI tools — that sent consumption into the tens of trillions of tokens. The pattern across large organizations is consistent: deploy the tools, lose control of the costs, then scramble to impose limits after the damage is done.
What does this mean for the business owner who cannot afford a dedicated AI team? It means the enterprise approach to AI — unlimited token budgets, thousands of engineers, months of integration — is not just expensive. It is structurally broken for anyone who does not have a billion-dollar operating budget and a team of financial controllers watching the meter.
Viktor works on a fundamentally different model. There is no token meter running in the background. No per-interaction billing that compounds with every reasoning step. You pay for credits, you use credits, and you see exactly what you spent. Viktor runs on Claude, GPT-4, and Gemini — all three included in one credit balance — and selects the right model automatically so you are not picking between tools or worrying about which API costs less per call.
More importantly, Viktor does tasks, not just answers. It acts inside your tools — writing, researching, building, managing workflows — so you get completed work rather than a coding suggestion you then have to implement, review, test, and deploy yourself. Uber needed 5,000 engineers using AI coding assistants. A business owner using Viktor needs a laptop and a clear brief.
You get $100 of free credits to begin — no credit card, no time limit, no commitment. Explore Viktor properly. Do real work. When you are ready to go further, $50 comes straight off your first bill.
Disclosure: Some links in this article are affiliate links. If you choose to get started with Viktor using the links provided, I may receive a commission — at no additional cost to you. I only recommend tools I use and believe in.