Together AI

Together AI provides a fast inference platform for open-source LLMs including Llama, Mistral, Qwen, and more. Known for competitive pricing, high throughput, and a wide selection of models. Ideal for developers who want to use open-source models without self-hosting.

Official Website Documentation

Endpoints

POST

Chat Completions

/v1/chat/completions

Chat with open-source models at high speed.

Streaming

Vision

Function Calling

Models

meta-llama/Llama-3.3-70B-Instruct-Turbometa-llama/Llama-3.1-8B-Instruct-Turbomistralai/Mixtral-8x22B-Instruct-v0.1Qwen/Qwen2.5-72B-Instruct-Turbodeepseek-ai/DeepSeek-V3

Pricing

Model	Input Price	Output Price	Context Window
Llama 3.3 70B Turbo	$0.88	$0.88	131.1K
Llama 3.1 8B Turbo	$0.18	$0.18	131.1K
Mixtral 8x22B	$0.90	$0.90	65.5K
DeepSeek V3	$0.88	$0.88	65.5K

Capabilities

Streaming

Function Calling

JSON Mode

Fine-tuning

Embeddings

Image Input

Try it in Playground

Test Together AI APIs directly in our interactive playground with your own API key.

Try it in Playground

Together AI

Endpoints

Chat Completions

Capabilities

Try it in Playground

Overview

Together AI

Endpoints

Chat Completions

Capabilities

Try it in Playground

Overview