OpenRouter
Catalog quote- Input
- $0.18
- Output
- $0.18
meta-llamaListed Apr 23, 2025
Llama-Guard-4-12B
Llama Guard 4 is a natively multimodal safety classifier with 12 billion parameters trained jointly on text and multiple images. Llama Guard 4 is a dense architecture pruned from the Llama 4 Scout…
Source description
Llama Guard 4 is a natively multimodal safety classifier with 12 billion parameters trained jointly on text and multiple images. Llama Guard 4 is a dense architecture pruned from the Llama 4 Scout pre-trained model and fine-tuned for content safety classification. Similar to previous versions, it can be used to classify content in both LLM inputs (prompt classification) and in LLM responses (response classification). It itself acts as an LLM: it generates text in its output that indicates whether a given prompt or response is safe or unsafe, and if unsafe, it also lists the content categories violated.
Context window163.8K
Max output16.4K
Parameters12B
Provider endpoints6
Benchmark results0
Measured performance
All leaderboards ↗Benchmark results
No published results in this index yet
Scores appear here when an evaluation matches this exact model.
Access & cost
6 quotesPrices across providers
Your token mix
USD · listed token ratesdeepinfra
via Hugging Face- Input
- $0.18
- Output
- $0.18
- Input
- $0.18
- Output
- $0.18
- Input
- $0.144
- Output
- $0.144
- Input
- $0.18
- Output
- $0.18
- Input
- $0.27
- Output
- $0.27
Model specifications
- Developer
- meta-llama
- Modality
- Language
- Context
- 163,840 tokens
- Maximum output
- 16,384 tokens
- Weights
- Open weights
- Provider listed
- Apr 23, 2025
- First discovered
- Oct 16, 2026
Exact identities
- openrouter
meta-llama/llama-guard-4-12b - huggingface
meta-llama/Llama-Guard-4-12B - deepinfra
meta-llama/Llama-Guard-4-12B
Sources
Updated Oct 17, 2026
More from meta-llama
View all ↗meta-llamaListed Apr 5
Meta: Llama 4 Maverick
Llama 4 Maverick 17B Instruct (128E) is a high-capacity multimodal language model from Meta, built on a mixture-of-experts (MoE) architecture with 128 experts and 17 billion active parameters per forward...
1M contextImage inputTool callingNo reported score
No reported score
meta-llamaListed Apr 2
Llama-4-Scout-17B-16E-Instruct
The Llama 4 collection of models are natively multimodal AI models that enable text and multimodal experiences. These models leverage a mixture-of-experts architecture to offer industry-leading performance…
1.3M context17BImage inputTool callingNo reported score
No reported score
meta-llamaListed Apr 1
Llama-4-Maverick-17B-128E-Instruct-FP8
1M context17BImage inputOpen weightsNo reported score
No reported score
meta-llamaListed Dec 6
Llama-3.3-70B-Instruct-Turbo
Llama 3.3-70B Turbo is a highly optimized version of the Llama 3.3-70B model, utilizing FP8 quantization to deliver significantly faster inference speeds with a minor trade-off in accuracy. The model is…
131.1K context70BNo reported score
No reported score