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.