OpenAI introduced GPT-5.2 and GPT-5.3 Codex, both trained on NVIDIA's infrastructure, showcasing significant performance gains in coding and reasoning tasks. The models achieve top scores on various industry benchmarks, reflecting advancements in AI training techniques. NVIDIA's powerful systems enable faster development cycles for AI applications.

NVIDIA has released the Nemotron ColEmbed V2 models, designed for efficient multimodal document retrieval. These models utilize a late-interaction embedding approach to improve accuracy in handling text, images, and structured visual data. They perform well on the ViDoRe V3 benchmark, making them suitable for applications like multimedia search engines and conversational AI.