Google caps Meta's Gemini capacity amid compute shortage

Category: Markets
Status: developing
Sources: CNBC, HN discussion
Summary: The Financial Times reported 2026-06-28 that Google limited Meta's use of its Gemini models after Meta sought more capacity than Google could supply, telling Meta around March it could not meet the full purchase. The shortfall disrupted and delayed some Meta AI projects, and Meta told staff to use AI tokens more efficiently. Other Google cloud customers faced capacity limits too, but Meta was hit hardest because its demand was especially high. Reuters and CNBC carried the FT report; Google and Meta did not confirm specifics.
Why it matters: A frontier-model provider rationing capacity to a major customer shows AI compute scarcity now constrains roadmaps even for the largest buyers, raising the cost of single-provider model dependence.
Follow-up: Watch for confirmation from Google or Meta, Meta's mitigation (in-house models or alternative providers), and whether quota limits widen to other Gemini API customers.