Skip to main content

DeepSeek V3, a state-of-the-art open model, is now available. Try it now!

Model Library

Llama 3.3 70B Instruct

This updated version of Llama 70B achieves state-of-the-art results in reasoning, math, and instruction following. It offers similar performance to Llama 3.1 405B but with significantly better speed and lower cost.

Try It Learn more
Meta logo

Models

Moa

4,096 Context

Serverless

Models

Models

Whisper V3

Serverless

Whisper V3 Turbo

Serverless