NVIDIA Nemotron 3 Nano Omni
A single efficient open model for unified video, audio, image, and text reasoning.
NVIDIA Nemotron 3 Nano Omni is a specific open multimodal model release in the Nemotron 3/Nemotron model family. NVIDIA describes it as a 30B-A3B hybrid MoE large language model that unifies video, audio, image, and text understanding for enterprise Q&A, summarization, transcription, document intelligence, GUI/agentic workflows, and audio-video reasoning, with text output and up to a 256K-token context window.
Pricing
Model profile · Current snapshot
Input / 1M
$0.075
Output / 1M
$0.30
Blended / 1M
$0.131
Output TPS
329
TTFT (s)
0.56
Model Intelligence
Context window
256,000 tokens
Arena ranking
15
Benchmarkable
Yes
Model level
release
Intelligence Index
15
Coding Index
13.8
GPQA
0.47
HLE
0.05
SciCode
0.28
IFBench
0.63
LCR
0.41
TerminalBench Hard
0.08
TAU2
0.45
Recent stories
0 linked stories
No linked stories yet.