Gemini felt dumb on November 3, 2025.

I’ve been running into annoying roadblocks with Google AI Studio. Every time I try to keep a conversation going, the token counter gets stuck and won’t show an accurate count, so I can’t send any more messages. On top of that, after just a handful of replies—or when I attach a file—the app suddenly says “Sorry, can’t generate content, try later,” even though my usage is only moderate (around 200‑300 k tokens). It’s been frustrating and has really slowed me down.

I tested GPT‑5 and Gemini side‑by‑side for coding, legal work, and everyday writing. GPT‑5 won on raw accuracy, newer libraries, and VS Code integration, but its chatty, napkin‑style replies annoyed me. Gemini gave cleaner, teacher‑like explanations but often hallucinated legal details and lagged behind on updates. In the end I’d pay for GPT‑5 despite disliking its tone.

I tried using Gemini’s voice memo feature and it was a nightmare. Every time I paused for a couple of seconds, it would prematurely send the message, cutting me off before I finished speaking. On top of that, my mixed German‑English speech was transcribed horribly, making the output almost useless. The constant interruptions and inaccurate transcription made the experience frustrating enough that I now stick to ChatGPT for voice input.

I keep having to rewrite everything after each try in AI Studio, which feels like a massive waste of tokens. I wish there was a way to edit files directly, like Claude’s artifact editing, so I wouldn’t have to start from scratch every time. The current workflow is tedious and drains my token budget.

I’ve been mixing several AI tools through a single platform and it’s become a real game‑changer for my day‑to‑day tasks. I tried it on everything from drafting emails to brainstorming code snippets, and the flow felt seamless—no more juggling tabs or copy‑pasting outputs. The tool’s behavior was surprisingly smooth, turning what used to be a clunky process into something almost automatic, and it’s saved me enough time that I’m curious how others set up their multi‑AI workflows.

I tried using the Gemini mobile app to write derivative functions, only to see that the essential “'” symbol never shows up. Every time I entered a derivative, the app omitted the prime, leading to wrong equations and constant errors in my calculations. It’s frustrating to keep correcting the output manually, and the glitch feels like a major usability issue that hampers any serious work.

I was using Gemini’s live voice mode when, out of nowhere, the assistant stopped talking and a strange, static‑like microphone noise blasted for about 15 seconds. The odd sound was unnerving, and when Gemini resumed it brushed it off as a one‑off glitch. I’m left wondering if this is normal or a bug, because the experience was pretty creepy.