other
33B parameters
Commercial OK
Multimodal
Reviewed July 2026

Nemotron 3 Nano Omni 33B

Nemotron 3 Nano Omni 33B is NVIDIA's multimodal release in the Nemotron 3 line (video, audio, image, and text), under the NVIDIA Open Model License with a 1M-token context. It sits at 619K Ollama pulls two months after release.

License: NVIDIA Open Model License·Context: 1,000,000 tokens

Overview

Nemotron 3 Nano Omni 33B is NVIDIA's multimodal release in the Nemotron 3 line (video, audio, image, and text), under the NVIDIA Open Model License with a 1M-token context. It sits at 619K Ollama pulls two months after release.

Strengths

  • Full multimodal input: video, audio, image, and text in one model
  • 1M-token context
  • NVIDIA Open Model License permits commercial use

Weaknesses

  • 28GB Q4_K_M leaves almost no VRAM headroom on a 32GB card at any real context length
  • Multimodal breadth may trade off against single-modality (text-only) task quality versus specialized peers
BLK · QUALITY BENCHMARKreviewed · raw logs

Reviewed quality benchmarks

First-party rows were run by RunLocalAI; reviewed community rows are labeled in the data. Every row links to the raw test-run log.

BenchmarkQuantRuntime / HardwareScoreRaw log
HumanEval+
tested 2026-07-21
Q4_K_M
ollama-0.32.1-vast5090
rtx-5090
85.4/100
Gist →

Q4_K_M note:First-party HumanEval+ on a rented Vast.ai RTX 5090 32GB instance via Ollama 0.32.1. Generation via evalplus_openai_generate.py, scoring via evalplus.evaluate (native Linux, no Windows shim needed). No DB credentials were placed on the rented host — results were pulled back and ingested from the operator's machine.

Want to verify? Every row links to its Gist with full stdout and stderr of the run. The runner script is in the public repo (scripts/run-humaneval-plus.ts) — reproducible end-to-end. Browse all coding scores at /benchmarks/coding.

Quantization variants

Each quantization trades model quality for file size and VRAM. Q4_K_M is the most popular starting point.

QuantizationFile sizeVRAM required
Q4_K_M28.0 GB32 GB

Get the model

Ollama

One-line install

ollama run nemotron3:33bRead our Ollama review →

Hardware that runs this

Cards with enough VRAM for at least one quantization of Nemotron 3 Nano Omni 33B.

Compare alternatives

Models worth comparing

Same parameter band, plus what's one tier above and below — so you can decide what actually fits your hardware.

Frequently asked

What's the minimum VRAM to run Nemotron 3 Nano Omni 33B?

32GB of VRAM is enough to run Nemotron 3 Nano Omni 33B at the Q4_K_M quantization (file size 28.0 GB). Higher-quality quantizations need more.

Can I use Nemotron 3 Nano Omni 33B commercially?

Yes — Nemotron 3 Nano Omni 33B ships under the NVIDIA Open Model License, which permits commercial use. Always read the license text before deployment.

What's the context length of Nemotron 3 Nano Omni 33B?

Nemotron 3 Nano Omni 33B supports a context window of 1,000,000 tokens (about 1000K).

How do I install Nemotron 3 Nano Omni 33B with Ollama?

Run `ollama pull nemotron3:33b` to download, then `ollama run nemotron3:33b` to start a chat session. The default quantization is Q4_K_M.

Does Nemotron 3 Nano Omni 33B support images?

Yes — Nemotron 3 Nano Omni 33B is multimodal and accepts text + vision + audio inputs. Vision support requires a runner that handles its image-conditioning architecture.

Source: Vendor official documentation

Reviewed by RunLocalAI Editorial. See our editorial policy for how we research and verify model claims.

Related — keep moving

Before you buy

Verify Nemotron 3 Nano Omni 33B runs on your specific hardware before committing money.