GET STARTED
One API. Your next idea.
Connect your app to AI models through one familiar API. This is your guide to building with Tokely.
A familiar way to build
Tokely provides an OpenAI-compatible API for text and images, speech synthesis and asynchronous media tasks. Use your existing SDK, choose a model, and send your first request. Your API key and account balance work across available models.
Start with Chat Completions or use Responses for models that support it. Each model page shows availability, supported endpoints, streaming support and current prices. See the Media API guide for images, speech and video.
Your connection details
Use this base URL in your SDK. Authenticate every request with a Tokely API key.
| Setting | Value |
|---|---|
| Base URL | https://api.tokely.me/v1 |
| Authentication | Authorization: Bearer <TOKELY_API_KEY> |
| Request format | application/json |
A first look
Create a key in your dashboard and make sure your account has enough balance before running your first request.
curl https://api.tokely.me/v1/chat/completions \
-H "Authorization: Bearer $TOKELY_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-5.4-mini",
"messages": [{"role": "user", "content": "Hello, Tokely!"}],
"max_tokens": 1024
}'Build with what is available
The public catalog shows currently enabled models. GET /v1/models returns enabled IDs; check individual model pages for compatible endpoints and capabilities.
Text, image generation, speech synthesis and asynchronous video/audio tasks are enabled for supported models. See /docs/media for endpoints and examples. Embeddings are available at /v1/embeddings for text-embedding-3-small, text-embedding-3-large and text-embedding-ada-002; see the Embeddings guide. Reranking is not offered.