Pricing
| Type | Price (per 1M) |
|---|---|
| Input tokens | $0.20 |
| Output tokens | $0.20 |
Capabilities
VisionMultimodalReasoningFunction CallingTool UseJSON ModeCode Execution
About Llama 3 Sonar Small 32K Online
The Llama 3 Sonar Small 32K Online model, developed by Perplexity AI, leverages the Mistral-7B base model to deliver real-time data processing from the internet. Notably, it features a 32,000-token context window for handling longer input sequences, enhancing its ability for comprehensive understanding and response generation. It offers robust performance in tasks like text summarization and question answering, despite being smaller and faster than the Llama 3 Sonar Large model. However, users have noted that its memory capacity might not fully utilize the 32k tokens, occasionally missing information from earlier in conversations, and it primarily supports the English language.
Model Specs
Released2024-05-05
Parameters8B
Context28K
ArchitectureDecoder Only
Knowledge cutoff2024-03