Gemini Pro subscribers report that Google silently routes their queries to lighter Flash models, especially after they cancel auto-renewal, to cut costs and serve the mass market. Users note that 3.6 Flash feels like an incremental evolution, while 3.1 Pro remains superior for reasoning.
Users now suspect silent model switching, not just strategic prioritization.
A growing number of consultants and analysts report preferring Anthropic's Claude over OpenAI's ChatGPT for tasks like long document analysis, strategy memos, and summarizing client calls. Users cite Claude's superior long-context handling and less marketing-oriented writing as key advantages, though ChatGPT remains the default for general-purpose use.
A growing number of consultants and analysts report preferring Anthropic's Claude over OpenAI's ChatGPT for tasks like long document analysis, strategy memos, and summarizing client calls. Users cite Claude's superior long-context handling and less marketing-oriented writing as key advantages, though ChatGPT remains the default for general-purpose use.
Multiple users report that Gemini's system prompt appears to forbid web search, causing the model to fabricate information and ignore explicit search requests. This raises reliability concerns for developers relying on Gemini for fact-checking and API documentation.
Multiple users report that Gemini's system prompt appears to forbid web search, causing the model to fabricate information and ignore explicit search requests. This raises reliability concerns for developers relying on Gemini for fact-checking and API documentation.
Multiple users report a recent degradation in Gemini's reliability, including errors, hallucinations, and failure to follow instructions, particularly in document-processing tasks. The issues appear to have started within the past week, affecting long-time users who previously had few problems.
Multiple users report a recent degradation in Gemini's reliability, including errors, hallucinations, and failure to follow instructions, particularly in document-processing tasks. The issues appear to have started within the past week, affecting long-time users who previously had few problems.
Google announced changes to Gemini Apps model availability on personal accounts taking effect October 9, 2026. Free users will be limited to Flash Lite, AI Plus subscribers get Flash added, and AI Pro/Ultra subscribers keep access to Pro plus Deep Think. Users can also select an effort level (low, medium, high) per model, trading answer thoroughness against usage limits.
Google announced changes to Gemini Apps model availability on personal accounts taking effect October 9, 2026. Free users will be limited to Flash Lite, AI Plus subscribers get Flash added, and AI Pro/Ultra subscribers keep access to Pro plus Deep Think. Users can also select an effort level (low, medium, high) per model, trading answer thoroughness against usage limits.
A Google email informed an annual Plus subscriber that their plan will be downgraded on November 13th, ending access to the Pro model. The subscriber says only the Pro model is reliable enough for their use case, forcing a choice between upgrading to a Pro subscription or losing half a year of paid value.
A Google email informed an annual Plus subscriber that their plan will be downgraded on November 13th, ending access to the Pro model. The subscriber says only the Pro model is reliable enough for their use case, forcing a choice between upgrading to a Pro subscription or losing half a year of paid value.