OcuSmart AI switches to GPT‑5.5

OcuSmart AI said on May 21 that it had switched its ophthalmology platform to GPT-5.5 Instant as the default model across its modules, according to a post from the company’s @OphthalmologyAI account on X. The update applies to a tool described elsewhere as a ChatGPT-based system for clinical, research and educational ophthalmology. The company’s post framed the change as a speed-oriented update for clinical use. OpenAI introduced GPT-5.5 Instant on May 5 as the new default model in ChatGPT. ### Which part of the product changed? The May 21 post said GPT-5.5 Instant is now the default model across OcuSmart AI’s modules. The company described the change as supporting faster clinical decision-making within the platform, based on the social post cited in the briefing. A separate public listing for OcuSmart describes the product as an artificial-intelligence framework for clinical, research and educational ophthalmology. An article in *The Ophthalmologist* published in January 2025 identified Alfredo Di Giovanni as the developer of OcuSmart and described the platform as integrating generative AI into everyday ophthalmic practice. ### What is GPT-5.5 Instant, and why would a vendor pick it? OpenAI said on May 5 that GPT-5.5 Instant became ChatGPT’s default model and was designed to be “smarter and more accurate” with clearer, more concise answers. The company said the model is intended for broad day-to-day use and reported improvements in factuality, image analysis, STEM questions and deciding when to use web search. (theophthalmologist.com) OpenAI also said GPT-5.5 Instant produced 52.5% fewer hallucinated claims than GPT-5.3 Instant on high-stakes prompts in medicine, law and finance, and 37.3% fewer inaccurate claims on conversations users had flagged for factual errors. Those figures came from OpenAI’s internal evaluations, according to the company’s release. ### How does OcuSmart fit into ophthalmology software? (openai.com) Alfredo Di Giovanni wrote in *The Ophthalmologist* in January 2025 that OcuSmart was built to move beyond static image classification toward longitudinal support, including predictive curves for glaucoma, macular disease and progressive myopia. In that article, he said generative AI could update disease forecasts at each patient visit and compare observed outcomes with predicted trajectories. (openai.com) The company’s public ChatGPT listing describes OcuSmart in broader terms, saying it is for professional use and supports clinical and research work in ophthalmology. The available public material reviewed for this story did not detail whether the GPT-5.5 Instant switch changes pricing, regulatory status or any image-analysis workflow. (theophthalmologist.com) ### Is this an FDA-cleared diagnostic change? No public document reviewed for this story said the May 21 update was an FDA clearance, label expansion or new diagnostic authorization. The available evidence is a company social-media announcement about the default language model used inside the platform. The distinction matters because ophthalmology software spans several categories, from workflow and documentation support to image analysis and autonomous diagnostics. (chatgpt.com) The materials reviewed here describe OcuSmart as a generative-AI support tool, but they do not provide a new regulatory filing tied to the GPT-5.5 Instant update. ### What should readers watch next? The next concrete marker is whether OcuSmart publishes product documentation describing which modules now default to GPT-5.5 Instant and whether users can still select other models. A second reference point is OpenAI’s own release material for GPT-5.5 Instant, published May 5, which remains the main technical source for the model’s claimed speed and accuracy changes. (openai.com) (theophthalmologist.com)