WebLLMLLMConfig

Defined in: src/providers/llm/webllm/WebLLMLLM.ts:92

Configuration for the WebLLM in-browser LLM provider.

Remarks

Unlike server-side providers, WebLLM needs no API key or proxy — everything runs client-side via WebGPU. The only required field is model.

Example

const config: WebLLMLLMConfig = {
  model: 'Llama-3.2-1B-Instruct-q4f16_1-MLC',
  stream: true,
  systemPrompt: 'You are a helpful assistant running locally.',
  onLoadProgress: ({ progress, text }) => {
    console.log(`Loading: ${Math.round(progress * 100)}% - ${text}`);
  },
};

See

• LLMProviderConfig for inherited base properties (temperature, maxTokens, systemPrompt, etc.).
• Available WebLLM models

Extends

• LLMProviderConfig

Properties