Use Jutusta inside your AI tool
Claude and other AI assistants can talk to Jutusta directly: read text aloud, turn recordings into text, summarize and translate. You stay in control — connecting takes a couple of minutes and needs no programming.
Connect your AI tool
Most clients only need the server URL — they discover the tools and run the OAuth sign-in themselves. Pick your client below for the exact steps.
- Open Claude and go to Settings → Connectors.
- Click “+” and choose “Add custom connector”.
- Paste the server URL above and click “Add”. Claude opens the Jutusta sign-in page; approve the permissions you want to grant.
- In a chat, open the “+” menu → Connectors and switch Jutusta on.
Each tool family has its own permission: tts (speech synthesis), asr (transcription), voices, summary, translate and denoise. Approve only the ones you want the agent to be able to spend credits on — tools whose permission you skipped will say so when called.
Already connected before the tools expanded? Disconnect and reconnect Jutusta in your client, so the sign-in can ask for the new permissions.
What can a connected tool do?
Once connected, just ask your assistant in plain language — it uses Jutusta's skills on your behalf:
- reads text aloud with a natural voice, including in Estonian;
- turns an audio or video recording into text;
- condenses a long recording into a short summary;
- translates text into another language;
- removes background noise from a recording.
The work is paid for with the credits on your Jutusta account — exactly as if you used Jutusta yourself. You can disconnect in your tool at any time.
How does it work?
You don't need this part to connect — it explains, in more technical terms, what happens under the hood.
Jutusta is a standard OAuth 2.1 / OIDC authorization server, so an AI agent can call the speech API on a user's behalf. There are two ways to get a bearer credential — both are presented the same way at call time.
- On a user's behalf (OAuth 2.1) — The agent registers a client (dynamic client registration), runs the authorization-code + PKCE flow, and the user signs in and approves the scopes on a consent screen. Synthesis and transcription spend the authorizing user's credits.
- As your own bot (API key) —
An API key (prefixed
sk_jt_) works without a sign-in flow and suits your own bot or script. Create one on the developer API page.
There is also a native MCP endpoint at /mcp (Streamable HTTP, JSON-RPC) exposing tools for speech synthesis, voices and voice cloning, transcription, summaries, translation and noise removal — guarded by the same bearer token, with one OAuth scope per family.
Agents can discover all of this without reading this page. Step-by-step instructions for an LLM live at /auth.md.
Are you a developer? API keys, the full reference and a try-it panel live on the developer API page.