Google has imposed usage caps on Meta's access to its Gemini generative AI models, reportedly because surging AI demand is straining its computational capacity.
June 2026 · AI Compute
Google Caps Meta's Gemini Usage as AI Compute Runs Short
A custom limit on one of its largest customers signals how scarce AI computing capacity has become — and how that scarcity is now reshaping deals between the biggest tech firms.
2×
Compute capacity Google aims to double every 6 months
~1,000×
Targeted capacity increase over 4–5 years
$93B
Projected capital spending to keep pace
Consumer Plan Limits
Compute-based usage multipliers by plan
Limits now count actual compute use — factoring prompt complexity, model, features and conversation length — across a 5-hour rolling window plus a weekly cap.
User complaints
Free tier "became nearly unusable"
Even paid users hit caps quickly
Fell back to weaker models after using high-performance ones
Google's easing measures
Per-prompt consumption caps
Failed requests no longer counted
More detailed usage breakdown shown
The bigger picture
The Meta cap is a symbolic case of compute scarcity reaching into deals between major players.
Google · Gemini
OpenAI · ChatGPT
Anthropic
All face usage limits
Continue reading The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in ✓ Signed in — this article isn’t included in your current plan.Unlocking the full article…