Together AI | The AI Native Cloud

# Together AI > Together AI is the AI Native Cloud — a platform for running, fine-tuning, and deploying > open-source and frontier AI models at scale. Developers use Together for serverless > inference, dedicated endpoints, GPU clusters, fine-tuning, and code sandboxes via an > OpenAI-compatible API. ## Getting Started - [Quickstart](https://docs.together.ai/docs/quickstart): Get your first API call working in minutes - [OpenAI Compatibility](https://docs.together.ai/docs/openai-api-compatibility): Drop-in replacement for OpenAI SDK - [Pricing](https://www.together.ai/pricing): Per-token and per-compute pricing for all products - [API Authentication](https://docs.together.ai/reference/authentication-1): How to authenticate API requests - [Introduction](https://docs.together.ai/docs/introduction): Platform overview and capabilities - [Integrations](https://docs.together.ai/docs/integrations): Connect Together AI with your existing tools - [Multiple API Keys](https://docs.together.ai/docs/multiple-api-keys): Manage API keys across teams and projects ## Products - [Serverless Inference](https://www.together.ai/serverless-inference): Pay-per-token access to 200+ open-source and frontier models - [Dedicated Model Inference](https://www.together.ai/dedicated-model-inference): Reserved model capacity for production workloads - [Dedicated Container Inference](https://www.together.ai/dedicated-container-inference): Deploy custom containers on dedicated infrastructure - [Batch Inference](https://www.together.ai/batch-inference): Async large-scale inference at lower cost - [Fine-tuning](https://www.together.ai/fine-tuning): Full fine-tuning and LoRA adapters for custom models - [Accelerated Compute / GPU Clusters](https://www.together.ai/accelerated-compute): On-demand H100/H200/GB200/B300 clusters for training and inference - [Sandbox](https://www.together.ai/sandbox): Secure code interpreter and execution environment - [Managed Storage](https://www.together.ai/managed-s