---
title: "When AI Costs More Than the Engineer"
description: "Anthropic spends 2.3x payroll on compute. Top software firms spend 0.4x. Three scenarios for where the rest of the market lands by 2029."
categories: ["AI","SaaS","startups"]
keywords: ["AI spend","Ramp AI Index","Anthropic compute","SaaS gross margin","engineer productivity","AI breakeven","software unit economics","token economics"]
ai_summary: "Anthropic spends 2.3x its payroll on compute — $515k per engineer per year at today's $224k fully-loaded salary. The top 1% of software companies spend $89k, the median $137. Three 2029 scenarios bracket how that gap closes."
date: 2026-06-29
lastmod: 2026-07-17
canonical_url: https://www.tomtunguz.com/ai-spend-breakeven-2029/
author: "Tomasz Tunguz"
---


Anthropic spends 2.3x its payroll on compute.[^goldman] With ~5,000 employees & roughly $10b in inference & training spend in 2026, that works out to about $2m of compute per employee per year against a likely all-in comp of $500k+.[^anthropic-econ]

The rest of the software market trails. The top 1% of companies spend $89k per engineer per year on AI, 40% of a fully-loaded $224k senior engineer salary[^levels].[^ramp] The median spends $137. That is the gap : 2.3x at the frontier, 0.4x at the top of the market, near zero at the median.

How close does the rest of the market get? Three scenarios bracket the answer.

{{< email_image src="mmpqck3bothkt4rr3jf7" alt="Line chart showing three scenarios for AI spend as percent of engineer salary through 2029, with the Bull case converging to the Anthropic benchmark of 230 percent" width="540" height="390" >}}

Bear (token deflation wins), Base (top-1% trajectory tapers), Bull (rest of market reaches Anthropic's ratio by 2029). Each scenario maps to an annual AI bill per engineer.[^method]

| Year | Bear | Base | Bull |
|------|-----:|-----:|-----:|
| 2026 | $90k (40%) | $90k (40%) | $90k (40%) |
| 2027 | $106k (45%) | $164k (70%) | $258k (110%) |
| 2028 | $118k (48%) | $259k (105%) | $444k (180%) |
| 2029 | $106k (41%) | $363k (140%) | $596k (230%) |

In the Bull case, the AI bill alone per engineer matches an entire median-SaaS employee's revenue contribution.[^saas] Anthropic & OpenAI already generate $14m & $6.5m in revenue per employee, the highest in the Forbes Global 2000.[^epoch]

The cost structure follows the revenue structure.

Bull drivers : frontier model prices hold as training costs plateau & demand outruns supply. Agentic workflows consume tokens at orders-of-magnitude higher rates than chat, with Goldman Sachs projecting a 24-fold rise in token consumption by 2030.[^goldman-agentic] If a rival ships features faster, the AI bill stops being optional.

Bear counterweights : token prices have fallen 10x per year for three years.[^tokens] Open-weight models close the quality gap at a fraction of the cost.[^deepseek] Companies that ration usage by role or workload bend the curve.

{{< email_image src="odtjqwtlphmuix5rxxi1" alt="A wooden seesaw with a small engineer & laptop lifted high on the left while a stack of server racks sinks down on the right" width="540" height="232" >}}

One of these scenarios will land closer to truth in 2029. Which one are you modeling for 2027?

[^method]: Methodology. Senior engineer fully-loaded comp anchors at $224k/yr today & grows ~5%/yr (BLS wage trend). Each scenario's % of salary path drives annual AI spend per engineer. Bear path (% of salary by year) : 40, 45, 48, 41. Base path : 40, 70, 105, 140. Bull path : 40, 110, 180, 230. Bear dollars rise through 2028 then dip in 2029 as the ratio falls faster than salary inflation.

[^ramp]: Ramp AI Index, June 2026. [ramp.com/data/ai-index-june-2026](https://ramp.com/data/ai-index-june-2026). Top-1% firms spend $7,449/employee/month ($89k/yr) on AI, growing 14.1% month-over-month; median firm spends $11.38/month ($137/yr); 680x spending gap between leaders & the median.
[^levels]: Senior software engineer fully-loaded comp anchor at $224k/yr blends Levels.fyi Q1 2026 base salary data with the U.S. Bureau of Labor Statistics Employer Costs for Employee Compensation 2026 benefits loading. Top-tier firms ride higher.
[^goldman]: Goldman Sachs, *The AI Economy in 2026*. At AI-native firms like Anthropic, compute spend runs ~2.3x staff costs, indicating a structural cost base where infrastructure dominates payroll. See also industry coverage : [valueaddvc.com/ai-spending](https://valueaddvc.com/ai-spending).
[^anthropic-econ]: Anthropic headcount ~5,000 per [SaaStr](https://www.saastr.com/anthropic-only-has-5000-employees-almost-no-one-has-ever-been-this-efficient-thats-by-choice/) (June 2026). Inference & training spend ~$10b in 2026 against ~$5b revenue, via [Fortune AI capex coverage](https://fortune.com/2026/04/30/big-tech-hyperscalers-will-spend-700-billion-on-ai-infrastructure-this-year-with-no-clear-end-in-sight-eye-on-ai/). $10b / 5,000 = $2m compute per employee. All-in comp at top AI labs runs $500k+ per [Levels.fyi Anthropic data](https://www.levels.fyi/companies/anthropic/salaries).
[^saas]: Public SaaS revenue-per-employee benchmarks from KeyBanc Capital Markets SaaS Survey & OPEXEngine 2025-26 cohorts. Median ~$250k; top-quartile $400k-600k depending on company stage & vertical.
[^epoch]: Epoch AI, *Revenue Per Employee at AI Companies*, 2026. [epoch.ai/data-insights/revenue-per-employee-ai-companies](https://epoch.ai/data-insights/revenue-per-employee-ai-companies). Anthropic ~$14m, OpenAI ~$6.5m per employee, the highest in the Forbes Global 2000.
[^goldman-agentic]: Goldman Sachs Research forecasts agentic AI workloads driving a 24x increase in token consumption by 2030 vs current chat-dominated usage patterns.
[^tokens]: OpenAI's GPT-4 class input pricing fell from $30 per million tokens at launch (March 2023) to under $3 by 2026, roughly a 10x per year deflation rate on equivalent capability. Similar declines visible across Anthropic Claude & Google Gemini SKUs.
[^deepseek]: DeepSeek-V3 & subsequent open-weight releases delivered frontier-comparable benchmarks at 1/10th to 1/30th the API cost of leading proprietary models, per Ramp's June 2026 observation that top firms are "mixing frontier models with cheap open-source" to control costs.