AI Hub
← All models

Llama Guard 4 12B vs Sonar Reasoning Pro

Meta vs Perplexity — benchmarks, pricing, and capabilities side by side.

  • Llama Guard 4 12B is cheaper ($0.18 vs $2.00 per 1M input)
  • Llama Guard 4 12B has a larger context window (164K)
Llama Guard 4 12BSonar Reasoning Pro
Intelligence index95.7
DeveloperMetaPerplexity
TypeMultimodalMultimodal
AccessOpen weightsAPI only
Context window163,840 tokens128,000 tokens
Input price$0.18 / 1M$2.00 / 1M
Output price$0.18 / 1M$8.00 / 1M
Speed
ReleasedApril 30, 2025March 7, 2025
Parameters
Input modalitiesImage, TextText, Image
Output modalitiesTextText