One reason I took a break from ChatGPT a few months ago (I’m back now) was how sick to death I got of its constant emojis, especially when it came to lists. The brain emoji was my least favorite, along with the green checkmarks, the pointy fingers, and the yellow “hazard” signs.
Well, I’ll believe it when I see it, but with its latest “instant” model, OpenAI promises that we’ll be getting way less of those “gratuitous” emojis in ChatGPT’s responses.
Available now for all ChatGPT users—I haven’t seen it on my account yet, so it must still be rolling out—GPT-5.5 Instant is now the default ChatGPT model, replacing GPT 5.3 Instant.
Among other qualities, OpenAI says GPT-5.5 Instant is more accurate, serving up 52.5 percent fewer hallucinations than GPT-5.3 Instant on “high-stakes” prompts relating to medicine, law, and finance, while cutting “inaccurate claims” by 37.3 percent on “especially challenging” chats. Those are encouraging figures, but we’ll have to wait for independent benchmarks to evaluate its real-world performance.
The new model should also offer “tighter and more to-the-point” answers “without losing substance,” OpenAI says. Perhaps most importantly, GPT-5.5 Instant “avoids things that can make responses feel cluttered, like gratuitous emojis.” Let’s hope so.
I’ve found that ChatGPT has been less emoji-happy in recent weeks, particularly when switching to the Pro model (which, currently, is GPT-5.5). Still, the odd smiley or thumbs-up has still managed to appear from time to time for the previous “instant” model, as opposed to Gemini and particularly Claude, which are practically emoji-free.
Other features in GPT-5.5 Instant include increased effectiveness at pulling in context from previous chats, files, and Gmail (provided you’ve enabled ChatGPT’s Gmail connector). The new model will “intelligently” decide whether to add personalization from other sources and is “faster” at scouring old chats, OpenAI says.
Finally, you’ll be able to review which “memory” sources were used in a given chat no matter which ChatGPT model you’re using, which is good for nixing a memory source that’s out of date.
I’ll have more to say once I’ve had a chance to give GPT-5.5 Instant a trial run, and hopefully we’ll have comprehensive benchmarks from independent testers soon. But if ChatGPT’s new “instant” model does manage to dial back the emojis, that’ll get a big smiley from me—plus maybe a green checkmark or two.
