Nvidia combines speech, vision, and text in new AI model

Nvidia launched Nemotron 3 Nano Omni, a compact multimodal AI model that handles text, speech, and visual inputs in one system.

The company says the design is aimed at better reasoning and context awareness for autonomous AI agents.

Nvidia also claims improved speed and accuracy, but those results still need independent verification.

The release highlights a broader shift toward smaller multimodal models that are easier to deploy in enterprise settings.