[Feature Request] <Small Sonar Model for fast response - Use a 7B to accelerate COT in my agents>

[YoannBuzenet]

🚀 Feature Request

Offering a 7B Llama 3 for really fast response.

🔍 Problem Statement

During COT for my agents, Sonar takes sometimes up to 15 seconds on a search to get the complete response, it is too long

💡 Proposed Solution

Offering Llama 7B would be great - I don't need perfect phrasing but a synthetic summary of the websearches.

📌 API Impact

New model offering

🔄 Alternatives Considered

I'm using Sonar currently, which is great but sometimes slow all my agents because I'm waiting for the promise to complete and it can be so long sometimes

📎 Additional Context

/

---

📧 For direct inquiries, please reach out to api@perplexity.ai