Overview
Connect your application to AI models through AudioLab’s model gateway and routing APIs.
Build with AudioLab
AudioLab Router connects your application to text, image, video, and audio models. Use workspace API keys to call enabled models, choose routing behavior, and track requests and usage in the dashboard.
Start with the quickstart to create a key and verify your connection. Then choose the guide for your application’s task. Model availability and supported options depend on your deployment; a catalog entry alone does not confirm that a model is enabled.
Choose a Task
| I Want To… | Guide |
|---|---|
| Generate text, maintain a conversation, or call tools | Text |
| Generate an image | Images |
| Generate and download a video | Video |
| Turn text into audio | Text to Speech |
| Transcribe a recording or add live captions | Speech to Text |
| Build a native audio conversation | Realtime Audio |
| Create vectors for search | Embeddings |
| Rank search results | Reranking |
| Extract text and layout from documents | OCR |
| Classify content | Moderation |
| Use a model with the native inference protocol | Native Inference |
Use Models and Providers to discover supported capabilities. Use Model Routing to understand direct model selection, automatic aliases, and named routers.
Understand the Request Model
A workspace owns API keys, credits, settings, and request history. An API key authenticates your application within that workspace. Its scopes grant access to specific operations, such as llm.write for language inference or audio.write for audio inference.
Each inference request selects a model ID and an endpoint. The model determines capabilities; the endpoint defines the request and response format. Your deployment’s base URL determines where the request goes. All model types and workspace operations share one public API URL: https://api.tryaudiolab.ai.
Keep these values together in your application’s configuration. Consult Authentication for credentials and the API reference for exact contracts.