Skip to content
RankShark

Nemotron 3.5 Content Safety

NVIDIA Nemotron 3.5 Content Safety is a compact 4B-parameter multimodal guardrail model from NVIDIA, fine-tuned from Google Gemma-3-4B.

ModelOpen sourceby NVIDIAsince 2026
31.5

SharkScore

#266 in Models

Why it scores 31.5

Nemotron 3.5 Content Safety has a SharkScore of 31.5/100, ranked #266 of 608 in Models (updated 2026-11-19).

Each part compares Nemotron 3.5 Content Safety with other listings in Models. Evidence covers 87% of the score's weight.

Adoption
8 · 30%
User reviews
no data · 25%
Maintenance
21 · 20%
Trust & transparency
80 · 15%
Momentum
no data · 10%

Paying never changes this score. How ranking works · method v2026-11-18.2

RankShark review of Nemotron 3.5 Content Safety

Automated evaluation by RankShark · 18 Nov 2026 · method v1.0 · How we review

We were not paid for this review.

As of 2026-11-18, Nemotron 3.5 is a text-and-image safety classifier with custom-policy support, priced at $0.20 per million input and output tokens on OpenRouter. (E1, MEASURED)

NVIDIA’s 4B-parameter model classifies prompts and optional model responses as safe or unsafe. It accepts text, an optional image, and an optional custom safety policy, and returns text labels. (E1)

Best for

  • Teams moderating text and single-image inputs or model responses. · model card
  • Developers who need moderation against custom safety policies. · model card
  • Teams deploying with Transformers, vLLM, or SGLang on documented Linux and NVIDIA GPU setups. · model card

Not for

  • Use cases requiring audio or video input; the model card documents text and image inputs. · model card, OpenRouter

Key features

  • Input and response moderation. The vendor says the model can classify a user prompt and an optional model response, returning safe or unsafe labels. · model card
  • Image input. The vendor says the model accepts text and an optional image; the model card describes support for text and a single image. · model card
  • Custom policies. The vendor says developers can provide a user-defined safety policy and receive a reasoning trace before the classification. · model card
  • Safety categories. The vendor says standard taxonomy mode can return the safety categories that were violated. · model card
  • Documented context length. The model card says up to 128K context; RankShark measured the OpenRouter listing at 131,072 tokens on 2026-11-18. · model card, OpenRouter
  • Deployment runtimes. The vendor lists Transformers, vLLM, and SGLang as runtime engines. · model card

Pros

  • Classifies both user prompts and optional model responses. · model card
  • Supports custom safety policies as well as standard taxonomy classification. · model card
  • The model card documents three runtime engines: Transformers, vLLM, and SGLang. · model card

Cons

  • The documented inputs are text and images, and the documented output is text; audio and video inputs are not listed. · model card, OpenRouter
  • The model card lists Linux and specific NVIDIA GPU hardware; it does not document other operating systems or hardware. · model card
  • No free plan or free trial stated on the pages we checked.
  • Whether customer data is used for model training is not stated on the pages we checked.

Pricing

PlanPriceLimits
OpenRouter per-token usage$0.20 per million input tokens; $0.20 per million output tokensNo monthly plan or included token allowance stated.

Checked 18 Nov 2026

Free plan
Not stated (checked 18 Nov 2026)
Free trial
Not stated (checked 18 Nov 2026)
Trains on customer data
Not stated (checked 18 Nov 2026)
Data retention
Not stated (checked 18 Nov 2026)
Data regions
Not stated (checked 18 Nov 2026)
Data processing agreement
Not stated (checked 18 Nov 2026)
AI models used
Not stated (checked 18 Nov 2026)
API
OpenRouter lists chat completions, completions, messages, and responses endpoints. · OpenRouter
SDKs
Not stated (checked 18 Nov 2026)
Webhooks
Not stated (checked 18 Nov 2026)
Platforms
The model checkpoint is hosted on Hugging Face, and an API listing is available on OpenRouter. · model card, OpenRouter
Security claims
No security certification stated (checked 18 Nov 2026)

What we measured

Context window (OpenRouter listing) · now
131,072 tokens
Price per million tokens (OpenRouter listing) · now
$0.2 in, $0.2 out
Package downloads · 30 days
34,694
How we tested · sources
  • Hourly uptime and response-time checks of the product's URL.
  • Repository, release and download data from GitHub, npm, PyPI and Hugging Face where they exist.
  • We did not sign up, pay, or use the product as a customer.
  • 18 Nov 2026: Plans or prices changed on the pricing page.

Something wrong? Report an error with a public link that shows it; corrections re-run the review.

Top alternatives to Nemotron 3.5 Content Safety

See all
  • 67.1

    Qwen3.8 27B is an open-weight dense vision-language model from Qwen.

    ModelOpen source
  • 65.5

    Qwen3.6-35B-A3B is an open-weight multimodal model from Alibaba Cloud with 35 billion total parameters and 3 billion active parameters per t

    ModelOpen source
  • 64.7

    Qwen3.8 Flash is a multimodal reasoning model from Alibaba.

    ModelOpen source
  • 64.3

    Qwen3.6 27B is a dense 27-billion-parameter language model from the Qwen Team at Alibaba, released in April 2026.

    ModelOpen source