llama-guard-4-12b

meta/llama-guard-4-12b

NVIDIA Build Card

Multi-modal model to classify safety for input prompts as well output responses.

Context Length

163,840 tokens

Input Pricing

$0.18 / 1M

Output Pricing

$0.18 / 1M

Recent Latency

No probes

Model Architecture & Limits

Canonical IDmeta/llama-guard-4-12b

Publisher / Originatormeta

Model FamilyLlama 3

ArchitectureTransformer-based (Dense feedforward early-fusion architecture, pruned from Llama 4 Scout by removing routed experts and router layers, retaining only the shared expert.)

Context Window Limit163,840 tokens

Max Output Tokens4,096 tokens

Endpoint Protocolopenai-chat

Capabilities & Modal Surfaces

Chat Conversation✕ No

Tool / Function Calling✕ No

Reasoning / Thinking✕ No

Structured Output (JSON)✕ No

Vision / Multimodal✓ Yes

Audio Processing✕ No

License & Restrictions

GOVERNING TERMS: The trial service is governed by the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf); and the use of this model is governed by the [NVIDIA Community Model License](https://www.nvidia.com/en

Commercial and research use permitted under governing model and API terms.