It’s not good news for those who rely on free AI tools to generate content, as major AI companies are adjusting the usage limits for their latest models. Over the past few days, Google and OpenAI have adjusted how much users can access Gemini 3 Pro and Sora 2 without paying. The move signals a shift in how major AI platforms plan to manage heavy demand and rising compute costs.
Google and OpenAI reduce free access to Gemini 3 Pro and Sora 2 AI tools.(AFP)
Google Reduces Free Prompts and Image Generations
Google made the first change. When Gemini 3 Pro launched last week, its support page clearly stated that free users could make “up to 5 prompts per day” and generate three images through Nano Banana Pro. Now the page no longer lists specific numbers. Instead, it says free access comes with “daily limits [that] may change frequently.”
For image generation, Nano Banana Pro shortly reduced its limit to two images a day before returning to three. Google added that these numbers could vary depending on demand.
Also read: iQOO 15 vs OnePlus 15 vs Realme GT 8 Pro: Which one offers the better overall package?
OpenAI Limits Video Generations on Sora 2
OpenAI followed a similar path with Sora 2. The company reduced the number of daily video generations for free users from 30 to six. Bill Peebles, who leads Sora at OpenAI, explained the change on X, saying the platform’s “GPUs are melting” due to high usage. The comment reflects the strain caused by video generation, which requires extensive computing power.
The latest adjustments from both companies suggest a strategy shift: free access remains available, but it will likely become more restrictive as demand grows. Google is direct in stating that paid users will get priority. Its support page notes that if capacity drops, limits for free accounts will tighten first to maintain service quality.
Also read: Incognito isn’t always invisible: How to truly delete your browsing history
Paid Plans Offer More Features
Paid plans offer significantly higher limits across prompts, image creation, context size, and advanced features. For instance, the context window rises from 32,000 tokens for free users to 1 million for Pro and Ultra subscribers, allowing much larger files and more detailed analysis. Video generation, deep research, agent capabilities, and higher image limits also stay behind paywalls.
Also read: Realme Pad 3 India launch timeline, colour and storage options tipped online – All details
The pattern shows that as generative AI evolves, companies are pushing users toward subscription tiers to support resource-heavy services. Free access remains an entry point, but the tools’ full capabilities now sit clearly behind monthly plans.
