Key Points
- OpenAI limits free Sora video generations to six per day.
- Google removes the five‑prompt daily cap for free Gemini 3 Pro users.
- Nano Banana Pro image generation is reduced to two images per day for free accounts.
- Paid tiers for both OpenAI and Google retain higher usage allowances.
- Both companies cite high GPU demand and cost as the reason for the cuts.
- The moves reflect a broader shift toward monetizing compute‑heavy AI services.
ChatGPT Sora
Free‑Tier Reductions from OpenAI and Google
In a coordinated move, OpenAI and Google have quietly lowered the amount of AI‑generated content that free users can create. OpenAI announced that the Sora video‑generation system will now allow only six video generations per day for users on the free plan. The change is framed as a response to “melting” GPUs, reflecting the intense compute load that video synthesis demands.
Google’s adjustment affects its Gemini 3 Pro model. The previous limit of five prompts per day for free users has been removed, and the platform now offers only “basic access” with daily limits that may vary. At the same time, the Nano Banana Pro image‑generation tool, linked to Gemini 3 Pro, is capped at two images per day for free accounts. Users on Google’s paid AI plans continue to receive higher allowances—up to 100 prompts per day on the Google AI Pro tier and up to 500 prompts per day on the Google AI Ultra tier.
Rationale Behind the Limits
Both companies point to surging demand and the associated cost of running powerful graphics processing units (GPUs). OpenAI’s spokesperson, Bill Peebles, highlighted that the company’s GPUs are operating at capacity, prompting the need to ration free usage so that more people can still access Sora. Google’s support article similarly notes that “basic access” may change frequently, implying an ongoing effort to balance load and cost.
Implications for Users and the AI Market
The simultaneous tightening of free access suggests a broader industry trend. As AI models become more compute‑intensive—especially for video and high‑resolution image generation—companies are increasingly nudging users toward paid tiers to offset operating expenses. While free users now face stricter caps, the core services remain available, and paid subscribers retain their existing limits.
Analysts see these adjustments as an early sign of a maturing AI ecosystem where the economics of large‑scale model deployment drive pricing strategies. The changes may encourage users who rely heavily on AI‑generated media to consider subscription plans, thereby aligning revenue with the high costs of GPU infrastructure.
Source: techradar.com