What is the Suki Dictation SDK?
What is the Suki Dictation SDK?
The Suki Dictation SDK is a JavaScript and React library for dictation in the browser. It lets you add speech-to-text to your web application using Suki’s hosted dictation experience, authentication helpers, and callbacks.Read the Dictation SDK Overview for supported packages, In-field and Scratchpad modes, and how a typical session flows.
What are In-field mode and Scratchpad mode?
What are In-field mode and Scratchpad mode?
You can build dictation in two ways:
- In-field: Dictation overlays a target field. This pattern fits notes, forms, and similar workflows.
- Scratchpad: Dictation uses a floating panel that is not tied to a single input.
What is the difference between the Dictation SDK and the Dictation APIs?
What is the difference between the Dictation SDK and the Dictation APIs?
Dictation SDK: Your app embeds dictation in the browser. The SDK handles SDP authentication with
SukiAuthManager, the iframe lifecycle, and callbacks such as onSubmit.Dictation APIs: You call Suki’s REST and WebSocket dictation APIs from your own code without using this iframe SDK. That path is documented under Audio dictation and the Dictation API reference.The Introduction page includes a short note that compares the two approaches.Which npm packages should I install?
Which npm packages should I install?
Every setup uses
@suki-sdk/core for SukiAuthManager. You then add one surface package:-
JavaScript (vanilla or non-React frameworks):
@suki-sdk/dictationplus@suki-sdk/core. -
React:
@suki-sdk/dictation-reactplus@suki-sdk/core. You still importDictationClientfrom@suki-sdk/dictationin examples when you construct the client in code.
Does the Dictation SDK run in Node.js or on the server?
Does the Dictation SDK run in Node.js or on the server?
No. The product is meant for browser embedding. Your page needs a real
HTMLIFrameElement, postMessage, and layout so the hosted UI can mount and size correctly.Not supported: Using Node.js as the runtime that hosts DictationClient or the dictation iframe, or relying on SSR alone to render dictation instead of initializing on the client after the DOM (and your container, when you use one) exists.Read the runtime section in Error handling and Prerequisites.