OpenAI on Thursday launched GPT-5.2, an updated version of the model that runs its ChatGPT chatbot and other AI tools.
It scores higher than previous models on OpenAI's tests of performance for code generation, recognizing charts and graphs, and factual accuracy.
"We designed GPT‑5.2 to unlock even more economic value for people; it’s better at creating spreadsheets, building presentations, writing code, perceiving images, understanding long contexts, using tools, and handling complex, multi-step projects," OpenAI writes.
OpenAI claims GPT-5.2 "performs at or above a human expert level" in certain tasks like preparing spreadsheets and presentations.
The new model is available as "Instant," "Thinking" and "Pro" versions, which OpenAI says can deliver more complex and high-quality results depending on the user's needs. The new model will roll out gradually, beginning with paid plans.
GPT-5.1 will remain available for several months as a legacy version.
RELATED STORY | Disney invests $1B in OpenAI in deal to bring characters like Mickey Mouse to Sora AI video tool
OpenAI's new model comes as competition in the AI market heats up.
Google in November released a new version of its Gemini AI model, which it claims is faster and more accurate than its previous models.
The company is also partnering with the U.S. Defense Department to provide its Gemini AI tools to the Pentagon, and plans to launch a new set of AI-infused eyeglasses for consumers in 2026.
Anthropic also released an updated version of its Claude Opus 4.5 model in November.
