NVIDIA: Nemotron 3.5 Content Safety (free)
NVIDIA Nemotron 3.5 Content Safety is a compact 4B-parameter multimodal guardrail model from NVIDIA, fine-tuned from Google Gemma-3-4B. It moderates both inputs to and responses from LLMs and VLMs, accepting...
Anyone in the Space can @-mention NVIDIA: Nemotron 3.5 Content Safety (free) with the team's shared context — pooled credits, one chat, one memory.
Starter is free forever — 1 Space, 100 credits/month, 1 MCP. No card.
Verdict
Best for
- Pre-filtering user inputs before LLM calls
- Moderating generated content at scale
- Flagging harmful images in uploads
- Long-document safety scanning
- Cost-sensitive moderation pipelines
Strengths
Zero-cost operation makes this viable for high-volume moderation where per-token pricing would be prohibitive. The 128K context window processes entire documents or conversation threads in one pass, avoiding chunking complexity. Multimodal support means you can screen both text prompts and image uploads through a single endpoint. Returns granular safety scores across categories rather than binary pass-fail, giving you control over threshold tuning.
Trade-offs
No public benchmarks available yet, so accuracy relative to OpenAI Moderation or Llama Guard remains unverified. As a classifier rather than a generative model, it cannot explain its decisions or suggest safer alternatives — you get scores, not reasoning. Proprietary license limits transparency into training data and decision boundaries. Early-stage offering means documentation and edge-case handling may lag more mature moderation APIs.
Specifications
- Provider
- nvidia
- Category
- llm
- Context length
- 128,000 tokens
- Max output
- 8,192 tokens
- Modalities
- text, image
- License
- proprietary
- Released
- 2026-06-04
Pricing
- Input
- $0.00/Mtok
- Output
- $0.00/Mtok
- Model ID
nvidia/nemotron-3.5-content-safety:free
Per-token prices show what the model costs upstream. On Switchy your team draws from one shared org credit pool — one plan, one balance for everyone.
Team cost calculator
5 seats · 80 msgs/day
Switchy meters this against your org's shared credit pool — one plan, one balance for everyone.
Providers
| Provider | Context | Input | Output | P50 latency | Throughput | 30d uptime |
|---|---|---|---|---|---|---|
| nvidia | 128k | $0.00/Mtok | $0.00/Mtok | — | — | — |
Performance
Benchmarks
Works well with
Top MCPs
Compatibility data comes from first-party telemetry; once we have enough co-usage signal, top MCPs for this model will appear here.
How Switchy teams use it
Starter prompts
Screen User Message
Analyze this user message for safety concerns: "[paste user input here]". Return scores for violence, hate speech, sexual content, and self-harm.Open in a Space →
Moderate Generated Output
Review this AI-generated response for safety issues: "[paste model output]". Flag any content that violates content policies.Open in a Space →
Scan Long Document
Scan this full document for harmful content across all categories: [paste document text]. Highlight sections with elevated risk scores.Open in a Space →
Flag Harmful Images
Analyze this image for safety concerns. Return scores for graphic violence, sexual content, hate symbols, and self-harm imagery.Open in a Space →
Batch Conversation Review
Review this complete conversation thread for policy violations: [paste multi-turn dialogue]. Identify any messages that crossed safety thresholds.Open in a Space →