Base44 Superagents-AI agent that does it all
Mini toolsOmniVinci
By NVIDIA
OmniVinci is an omni-modal transformer designed for unified perception across visual, auditory, and textual streams. It introduces new architectural and data-curation strategies to efficiently train a 9B-parameter model on roughly 0.2 trillion tokens while still reaching state-of-the-art performance on multi-modal benchmarks. The system supports tasks like video question answering, audio understanding, captioning, and cross-modal retrieval, targeting use cases where a single model needs to reason over complex, time-varying inputs.
New
Image
Gen 4
Overview
OmniVinci is NVIDIA’s 9B omni-modal LLM that jointly understands images, video, audio, and text, achieving strong cross-modal reasoning with only about 0.2T training tokens.
About NVIDIA
Industry:
Computer Hardware Manufacturing
Company Size:
36000
Location:
Santa Clara, California, US
Website:
nvidia.com
Tools using OmniVinci
No tools found for this model yet.
Last updated: February 25, 2026
