I tested OpenAI’s three claims about GPT-5.5 Instant, and only one fully held up

OpenAI replaced GPT-5.3 Instant with GPT-5.5 Instant as ChatGPT's default model, but only one of three claims about the new model held up. The claims were likely made to promote the new model. Engineers should be aware of the potential limitations of GPT-5.5 Instant. Further testing is needed to fully understand the capabilities of the new model. Users should be cautious when relying on the new model.