API Reference

Provider Functions

webLLM(modelId, settings?)

Creates a WebLLM model instance.

Parameters:

• modelId: The model identifier from the supported list of models
• settings (optional): Configuration options
  • appConfig?: AppConfig - Custom app configuration for WebLLM
  • initProgressCallback?: (progress: number) => void - Progress callback for model initialization
  • engineConfig?: MLCEngineConfig - Engine configuration options
  • worker?: Worker - A web worker instance to run the model in for better performance

Returns: WebLLMLanguageModel

webLLM.embeddingModel(modelId, settings?)

Creates a WebLLM embedding model instance.

Parameters:

• modelId: The embedding model identifier
• settings (optional): Configuration options
  • appConfig?: AppConfig - Custom app configuration for WebLLM
  • initProgressCallback?: (progress: WebLLMProgress) => void - Progress callback for model initialization
  • engineConfig?: MLCEngineConfig - Engine configuration options
  • worker?: Worker - A web worker instance to run the model in
  • maxEmbeddingsPerCall?: number - Maximum texts per call (default: 100)

Returns: WebLLMEmbeddingModel

Available Models:

See WebLLM config for the latest list of supported models.

Utility Functions

doesBrowserSupportWebLLM()

Quick check if the browser supports WebLLM. Useful for component-level decisions and feature flags.

Returns: boolean - true if browser supports WebGPU, false otherwise

Model Methods

WebLLMLanguageModel.availability()

Checks the current availability status of the WebLLM model.

Returns: Promise<"unavailable" | "downloadable" | "available">

WebLLMLanguageModel.createSessionWithProgress(onProgress?)

Creates a language model session with optional download progress monitoring.

Parameters:

• onProgress?: (progress: number) => void - Optional callback that receives progress values from 0 to 1 during model download

Returns: Promise<WebLLMLanguageModel> - The configured language model instance

WebLLMLanguageModel.isModelInitialized

Property that indicates if the model is initialized and ready to use.

Returns: boolean

Worker Handler

WebWorkerMLCEngineHandler

Re-exported from @mlc-ai/web-llm for Web Worker usage.

import { WebWorkerMLCEngineHandler } from "@browser-ai/web-llm";

const handler = new WebWorkerMLCEngineHandler();
self.onmessage = (msg: MessageEvent) => handler.onmessage(msg);

Types

WebLLMUIMessage

Extended UI message type for use with the useChat hook that includes custom data parts for WebLLM functionality.

type WebLLMUIMessage = UIMessage<
  never,
  {
    modelDownloadProgress: {
      status: "downloading" | "complete" | "error";
      progress?: number;
      message: string;
    };
    notification: {
      message: string;
      level: "info" | "warning" | "error";
    };
  }
>;

DownloadProgressCallback

The callback type for receiving model download progress. Shared across all @browser-ai packages.

type DownloadProgressCallback = (progress: number) => void;

WebLLMModelId

Type alias for model identifiers.

type WebLLMModelId = string;

WebLLMSettings

Configuration options for the WebLLM model.

interface WebLLMSettings {
  appConfig?: AppConfig;
  initProgressCallback?: (progress: number) => void;
  engineConfig?: MLCEngineConfig;
  worker?: Worker;
}