kayetanaOne API for every task.
Everything you try in the Playground is also available to your own apps. The REST API runs on your server, protected by API keys with their own monthly limits.
What the API offers under /v2.
| Endpoint | Task |
|---|---|
| chat | Chat, including documents and sources, streaming with server-sent events |
| embed, rerank | Embed text and sort it by relevance |
| classify | Assign labels to text |
| tokenize, detokenize | Split text into tokens and back |
| transcribe | Turn speech recordings into text |
| parse | Read documents |
| models | List installed models |
| datasets, embed-jobs | Manage and embed datasets |
| finetuning | Train your own models and classifiers |
One key per app, one limit per key.
Create a separate key for each app, like contract review, website search or your support bot. Every key gets a monthly limit. Once it's reached, the API returns status 429 instead of quietly running up usage.
- Bearer key authentication
- Monthly limit per key
- Requests and tokens per key at a glance
What makes the work easier.
Streaming
Chat responses arrive word by word via server-sent events.
Sources included
With the documents parameter, chat returns the answer along with citations.
Sample code
The Playground shows every request as code in Python, TypeScript and curl.
Bearer keys
Every request authenticates with its own key, one per app.
Usage per key
Tokens and requests per day, key and model, plus response time and error rate.
On your network
The API runs on your server, over HTTPS with your own certificate.
From the Playground to your own app.
Test in the Playground
Adjust the model and settings until the result looks right.
Copy the code
Click “Code” and copy the request in Python, TypeScript or curl.
Create a key
Create an API key for the app and set a monthly limit.
Watch usage
The usage page shows how many tokens the app consumes.
Build AI into your own tools.
In a demo, we'll walk you through the API and the right sample code for your use case.
Frequently asked questions
How do I authenticate?
With an API key as a bearer token. You create keys in the web interface.
What happens when a key hits its limit?
The API returns status 429. That keeps load predictable, and you can see right away which app is affected.
Does the API support streaming?
Yes. Chat responses stream via server-sent events, including answers with sources.
Is there sample code?
Yes. The Playground shows working code in Python, TypeScript and curl for every request.
Does the API cost extra?
No. The API is part of the platform, with no per-token or per-request billing.