Google reduces Gemini quotas: fewer free AI responses

What happened?

Google has quietly updated the quota system for its AI assistant Gemini, replacing the conversation-based limit with a token-based limit. According to Wired, free users now have a cap of 60 requests per hour, but each request consumes a variable number of tokens depending on the length of the prompt and response. In practice, this means that lengthy conversations or complex queries quickly exhaust the quota, drastically reducing the number of possible interactions. Previously, users could maintain multiple conversations without a clear token limit; the new system is more restrictive and opaque, as Google does not provide a public tool to monitor token consumption in real time.

This change represents a significant shift in Google's strategy. Since the launch of Gemini in December 2023, the company had promoted generous free access to compete with OpenAI's ChatGPT, which at the time had a limit of 50 messages every 3 hours. However, as adoption grew and inference costs soared, Google opted for a more restrictive approach. Wired reports that the new system was implemented without prior notice, causing confusion among users who noticed abrupt cuts in their conversations. The lack of transparency is particularly problematic: unlike OpenAI, which displays the token counter in the interface, Google does not offer real-time metrics, leaving users without planning capability.

Why is it important?

This change affects Gemini's free user base, which had grown rapidly since its launch. According to Similarweb data, Gemini recorded over 100 million monthly visits in March 2024, although the exact number of active users has not been disclosed. The quota reduction can be interpreted as a strategy by Google to boost paid subscriptions (Gemini Advanced, which costs $19.99 per month) or to manage infrastructure costs. However, it generates discontent among users who rely on the free version for everyday tasks such as writing, programming, or studying.

Moreover, the lack of transparency in token measurement makes it difficult for users to plan their usage and can lead to unexpected interruptions. This contrasts with competitors like ChatGPT, which offers a clear hourly message limit without depending on length, or Claude, which sets a daily limit of 100 messages. Google's approach is more similar to paid APIs, where the cost per token is explicit, but in the free realm it is confusing. This change could also have regulatory implications: in the European Union, the Digital Services Act requires transparency in access limitation practices, and Google's opacity could attract scrutiny from authorities.

Historically, Google has used quota changes as a tool to steer user behavior. For example, in 2022, Google reduced the free storage limit for Google Photos high-quality uploads, driving subscriptions to Google One. In the case of Gemini, the strategy is similar: making free usage less attractive to convert users into paying customers. However, the risk is that users simply migrate to more generous alternatives like ChatGPT or Claude, which have maintained stable limits.

What consequences will it have?

• Migration of free users: It is likely that a significant portion of Gemini's free users will explore alternatives. ChatGPT offers 50 messages every 3 hours (approximately 400 messages per day) without token limitation, while Claude allows 100 messages daily. If Google does not adjust its quotas, it could lose ground in the AI assistant market, especially among students and developers who depend on free access.
• Pressure on Google: The negative reaction on social media and forums like Reddit has been immediate. Users report that conversations of more than 10 exchanges exhaust the limit, making Gemini practically unusable for complex tasks. Google may be forced to clarify the rules or increase free quotas, as happened in 2023 when ChatGPT temporarily reduced its limit and then restored it after complaints.
• Acceleration of paid subscriptions: For users who already value Gemini for its integration with the Google ecosystem (Gmail, Drive, Maps), subscribing to Gemini Advanced ($19.99/month) could be an attractive option. This plan offers much higher limits (up to 1,000 requests per hour) and access to more powerful models. However, the price is high compared to ChatGPT Plus ($20/month) or Claude Pro ($20/month), which also offer similar benefits.
• Increased regulatory scrutiny: The European Commission has already shown interest in the practices of big tech companies in AI. The lack of transparency in Gemini's quotas could be investigated under the General Data Protection Regulation (GDPR) or the Digital Services Act, especially if it is shown that the opacity harms consumers. In the United States, the FTC could also intervene if it is considered a deceptive practice.

What should readers know?

If you are a free Gemini user, keep in mind that long queries consume more tokens. To optimize your usage, keep prompts concise and avoid lengthy conversations. Also, consider alternatives like ChatGPT (free with a limit of 50 messages every 3 hours) or Claude (100 messages per day on its free plan). Google has not announced changes to Gemini Advanced, which offers much higher limits.

For developers, the Gemini API remains a viable option with per-token pricing, although the change in the free interface could be an indication of future restrictions on the API. It is advisable to monitor Google's updates and adjust applications accordingly.

“Google's change is a reminder that free AI services are not sustainable in the long term; companies will eventually seek to monetize or restrict access.” — Analyst at TheVortiq.

In summary, this move by Google reflects a broader industry trend: the transition from generous freemium models to more restrictive ones as computing costs soar. The key for users will be to assess whether Gemini's value justifies a subscription or whether it is better to migrate to alternatives that offer greater transparency and clearer limits.