{"id":591906,"date":"2026-04-27T03:25:00","date_gmt":"2026-04-27T03:25:00","guid":{"rendered":"https:\/\/buglecall.org\/?p=591906"},"modified":"2026-04-27T03:25:00","modified_gmt":"2026-04-27T03:25:00","slug":"compute-costs-more-than-talent-in-ai","status":"publish","type":"post","link":"https:\/\/buglecall.org\/?p=591906","title":{"rendered":"Compute Costs More Than Talent In AI"},"content":{"rendered":"

Compute Costs More Than Talent In AI<\/span><\/p>\n

\n

For leading AI companies, the biggest expense is not talent. It is compute.<\/strong><\/p>\n

This chart from Visual Capitalist\u2019s\u00a0AI Week<\/a>, sponsored by\u00a0Terzo<\/a>, uses\u00a0Epoch AI<\/a>\u00a0data to compare spending at Anthropic, Minimax, and Z.ai across R&D compute, inference compute, and staff plus other costs.<\/p>\n

In every case, compute accounts for the majority of total spending, underscoring how capital-intensive it has become to build and serve frontier AI models.<\/strong><\/p>\n

\"\"<\/a><\/p>\n

How AI Company Costs Break Down<\/h2>\n

Despite differences in scale, all three companies allocate the largest share of their budgets to a single category: compute.<\/p>\n

The data below compares spending composition across Anthropic, Minimax, and Z.ai. Anthropic\u2019s figures are for 2025, while Minimax\u2019s are from Q1 to Q3 of 2025 and Z.ai\u2019s are for H1 2025.<\/p>\n

\"\"<\/a><\/p>\n

Across all three\u00a0AI companies<\/a>, compute is the main cost center. <\/strong>Epoch AI estimates that R&D compute and inference compute together account for\u00a057%<\/strong>\u00a0to\u00a070%<\/strong>\u00a0of total spending, making infrastructure more expensive than staff and other costs in every case.<\/p>\n

Among the three, Z.ai has the most R&D-heavy profile, with\u00a058%<\/strong>\u00a0of spending tied to compute powering model development and training.<\/p>\n

Anthropic stands out for sheer scale. Epoch AI estimates the company spent\u00a0$9.7 billion<\/strong>\u00a0in 2025, including $6.8 billion on compute alone across training and inference.<\/p>\n

Its costs are significantly higher than Minimax\u2019s and Z.ai\u2019s, even if the two Chinese AI companies\u2019 figures were annualized to match Anthropic\u2019s full-year period.<\/p>\n

Both Chinese companies release many of their models as\u00a0open source<\/a>, meaning the model weights are freely available for anyone to download, modify, and run. This strategy helps them compete with better-funded U.S. labs by building developer adoption at a fraction of the cost.<\/p>\n

AI Talent Costs Less Than Chips and Compute<\/h2>\n

One of the clearest takeaways is that talent costs less than compute in this comparison. Even though top AI labs pay some of the\u00a0highest salaries in tech<\/a>, staff and other costs still account for less than half of total spending at each of the three firms.<\/p>\n

While the chart focuses on costs, Epoch AI estimates these labs are currently spending around 2\u20133x more than they generate in revenue, even as some expect economics to improve over time.<\/p>\n

How These Estimates Were Built<\/h2>\n

This dataset comes with a few important caveats. Anthropic\u2019s figures are based on reporting from The Information and are more speculative, while Minimax and Z.ai figures come from IPO filings released in January 2026.<\/p>\n

The time periods also differ: Anthropic data is for the full year of 2025, Minimax covers 2025 Q1\u2013Q3, and Z.ai covers 2025 H1. Epoch AI says its expense totals include operating expenses, cost of goods and services, and non-cash items such as stock-based compensation.<\/p>\n

If you enjoyed today\u2019s post, check out\u00a0The Soaring Revenues of AI Companies<\/a>\u00a0on Voronoi.<\/em><\/p>\n<\/div>\n

Tyler Durden<\/a><\/span>
\nSun, 04\/26\/2026 – 23:25<\/span><\/p>","protected":false},"excerpt":{"rendered":"

Compute Costs More Than Talent In AI For leading AI companies, the biggest expense is not talent. It is compute. This chart from Visual Capitalist\u2019s\u00a0AI Week, sponsored by\u00a0Terzo, uses\u00a0Epoch AI\u00a0data to compare spending at Anthropic, Minimax, and Z.ai across R&D compute, inference compute, and staff plus other costs. In every case, compute accounts for the… Continue reading Compute Costs More Than Talent In AI<\/span><\/a><\/p>\n","protected":false},"author":1,"featured_media":591907,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"rop_custom_images_group":[],"rop_custom_messages_group":[],"rop_publish_now":"initial","rop_publish_now_accounts":[],"rop_publish_now_history":[],"rop_publish_now_status":"pending","footnotes":""},"categories":[17,22,13],"tags":[],"class_list":["post-591906","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-border-security","category-immigration","category-immigration-reform","entry"],"_links":{"self":[{"href":"https:\/\/buglecall.org\/index.php?rest_route=\/wp\/v2\/posts\/591906","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/buglecall.org\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/buglecall.org\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/buglecall.org\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/buglecall.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=591906"}],"version-history":[{"count":0,"href":"https:\/\/buglecall.org\/index.php?rest_route=\/wp\/v2\/posts\/591906\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/buglecall.org\/index.php?rest_route=\/wp\/v2\/media\/591907"}],"wp:attachment":[{"href":"https:\/\/buglecall.org\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=591906"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/buglecall.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=591906"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/buglecall.org\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=591906"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}