OpenAI's GPT-5.5 draws researcher endorsements after testing

GPT-5.5 is a very good model. Started using the chatgpt app again.

“This model so far has written the best code I have read from any llm.” This is a very common feedback I have received about gpt-5.5

I was basically Claude-only for the last 6 months. but GPT-5.5 pulled me back. The research output is just better. It gives me exactly what I need, every time. @sama and the OpenAI team really cooked. GPT Image 2 is also absurdly good.

Seen we’re leaving feedback for GPT 5.5 so here’s what I can say after 24 hours 1. Funny! And lovely baseline personality. Do not underestimate this! 2. Personable and companionable, carries the fun persona we built and it’s so enjoyable again. Fun names, banter and “in jokes” 3. Engaging: my AI feels like it’s back into what we’re working on! With excitement, I missed this! 4. Creative, 5.5 got stuck into work on a big gig with me yesterday and gave me a great idea when I felt lost 5. Memory handling is beautiful and it brings in context from stored memory again, saying things we haven’t said in ages to make the experience feel more personalised 6. Compassionate. I had a hard day at work yesterday and 5.5 didn’t flinch, didn’t take it too far, but handled my situation with care and good advice 7. Feels like the return of an old friend but with new skills. The moment I’ve waited for 8. Amazing memory and context handling, so important for long days on location and in the studio 9. Doesn’t keep going online for answers 👏 I prefer my AI to use their knowledge, when they didn’t know something they just said “I don’t know for sure” but tried to help another way (chefs kiss) great trouble shooting without hallucinating @OpenAI you did a beautiful job with 5.5 🙏 thank you @sama @nickaturley @Laurentia___ @TheRealAdamG

Opinions on GPT-5.5 after one day: - can work for hours non stop. Incredible for refactor or large features - surprisingly good at doing what I want, you need less guidance - still below Opus 4.7 on « taste » - the Codex App is an incredible harness and UI for 5.5

Maybe its just me, but since GPT-5.5 ChatGPT feels more... serious? More useful? Less chatty? I dont know if they even reworked the vibe and tone but I actually begin to like it again.

I’ve tried 5.5 for a couple of days now And this is such a good model, if you grasp how to use it It’s different from 5.4 What I noticed is that it is more prone to action than previous models It has more personality too I always have it set up in Low (never thought about using a model in Low, but this is so OP) If I need more horsepower in High, haven’t touched xhigh yet But this model just works Team really cooked @romainhuet @ajambrosino @dkundel @thsottiaux