They developed a think button are you shitting me

OpenAI says the new GPT-5.6 Luna is also more reliable with facts. In an internal evaluation involving financial, medical and legal questions, the company found that responses containing at least one factual error were 62 per cent less common with GPT-5.6 Luna compared with GPT-5.5 Instant.

So better than their cheapest shittiest model. Woo great accomplishment. Nothing burger of a statement there.

Good to see them struggling to stay relevant with vaporware of features.

If you're not paying, you're the product.

Read that again, giving ChatGPT 'unlimited access to free users'.

This is for their IPO.

Gotta hype and pump.

...And FYI, the spectacular drop after the IPO will be coming out of y'all's retirement accounts.

They can't even afford their current paying user base... Aren't they months from insolvency at their burn rate?

That can't be good for their already worrisome finances

That’s sort of what they did? Luna seems to be mostly on par with gpt-5.4, which was SOTA half a year ago

Really have to be desperate for Puerto use their slop if they are doing it.

Altman is full of shit Everything he says is a lie.

DeepSeek just released a model that is much cheaper to run and performs just as well as most GPT and Claude models. This is their reaction to try limit the damage from DeepSeek.

Trying so hard to get people dependent on their AI so they will pay for it later.