GPT 6.2 Astra had made the AI community wait a long time for its release, but many of the early testers don’t seem particularly enthused about what OpenAI has dished out.
OpenAI’s GPT 6.2 Astra has received some unusually harsh reviews from early testers. Wharton Professor Ethan Mollick, who tests out most models prior to their official release, said he found the model “uneven and weird”. “I’ve been using GPT-6.2 Astra prior to launch & I found it very uneven & weird. On one hand, it was harder to steer & work with than other models. On the other, when given an ambitious project (make a space elevator ride from scratch with Blender & Unity) it could deliver well,” he posted on X.
YouTuber Theo, who has had some scathing reviews of Google’s models recently, called it OpenAI’s worst model. “6.2 Astra is here. I’ve been testing it for awhile. I think this is the worst model OpenAI has ever released. I’ll share more soon,” he said on X. He’s even published a detailed post on why GPT 6.2 Astra didn’t quite work for him.
Every CEO Dan Shipper said his company had been testing the model for a while, but said that the models weren’t great, and a disappointing release compared to Fable 5.2. “We’ve been testing them for the last few weeks and, unfortunately, these models aren’t great. Overall a disappointing release compared to Fable 5.2. – Writing. We found Astra 6.2 to be a step back from Astra 6 on writing. 6.2 appears to have been trained to very extremely concise and actionable—so much so that its writing is hard to understand, and it gives short answers when long answers would be appropriate. – Coding. We found performance to be similar to Astra 6 on most of our benchmarks. However, in real coding tasks Astra 6.2 seemed to stop too early frequently, which made it quite frustating to use on end-to-end builds. – Design. Astra 6.2 has good default design instincts. But it struggles to vary its designs, resulting in output that has a noticeable “house style”,” he said on X.
Investor Matt Schumer was even more blunt. “I will not be writing a full review of GPT-6.2 Astra / Sol. I had a terrible experience testing them. They were essentially unusable. I do not recommend these models,” he posted on X.
It’s unusual for reviewers to post such scathing reviews of a top frontier model. GPT 6.2, surprisingly, has done reasonably well on benchmarks. It’s gone past Gemini 3 Carbon on the Artificial Analysis Intelligence Index with a score of 62, compared to Carbon’s 60, and is just 5 3 points behind Fable 5.2. OpenAI has also released some impressive benchmark results, which show the model beating Fable 5.2 and Gemini 3 Carbon on some measures. But some early testers seem to believe that these scores aren’t quite translating into how they found their experience with GPT 6.2 Astra to be like.