Technology20.05.2026

Google Gemini Post-Update: Users Complain of Performance Degradation

Now, complex tasks — such as video and image processing, deep data analysis, or simply lengthy sessions — exhaust the allocated limit extremely quickly.

On May 19–20, 2026, immediately following the Google I/O presentation and the release of the Gemini 3.5 Flash model, social media platforms and specialized forums saw a sharp spike in complaints regarding the AI’s performance. Many users note that the neural network has become less user-friendly and predictable in daily use.

The primary source of dissatisfaction stems from the system’s strict new limits and a significant drop in quality when in fallback mode. Demanding tasks—such as video and image processing, deep data analysis, or simply lengthy sessions—now exhaust the allocated quota extremely fast, blocking full access for hours. Once the limits are hit, the model automatically switches to a “Lite” version, which users found to be substantially weaker than the previous baseline, describing it as practically unusable for serious work.

An updated web design for Gemini drew an additional wave of criticism, with many complaining about excessive brightness, a “pretentious” look, and poor readability of the interface. At the same time, serious grievances targeted the actual quality of text generation. According to users, the AI’s responses have become more formulaic, “flat,” and oversaturated with clichés and emojis, while its precision in adhering to complex instructions and specific writing styles has noticeably declined. This has created numerous hurdles for those relying on the neural network to write articles, create posts, and edit materials. Combined with these issues, the audience has developed a general sense of system degradation, noting that the AI has become worse at maintaining context in long conversations.

However, the community’s reaction is not exclusively negative. Some users highly praise the 3.5 Flash architecture itself, pointing out a significantly increased processing speed and clear improvements in simple, everyday scenarios, as well as in writing code.

Such surges of discontent following major updates to Google products have become almost regular. The company traditionally attempts to strike a difficult balance between generation speed, server maintenance costs, and overall response quality. Nevertheless, many experienced users emphasize that in recent months, this trade-off has increasingly shifted toward performance restrictions and simplified logic. At the time of publication, Google had not issued any official comments regarding the latest wave of feedback.

Other news from the world of artificial intelligence