import * as webllm from "@mlc-ai/web-llm"; //you can also use the jsdelivr cdn or local mlc webllm npm files

const appConfig = {
  model_list: [{
    model: "https://huggingface.co/Aadeshisdoingsomething/gemma-3-270m-it-q4f16_1-mlc",
    model_id: "gemma-3-270m-it-q4f16_1-mlc",
    model_lib: "https://huggingface.co/Aadeshisdoingsomething/gemma-3-270m-it-q4f16_1-mlc/resolve/main/gemma-3-270m-it-q4f16_1-webgpu.wasm",
    required_features: ["shader-f16"]
  }]
};

const engine = await webllm.CreateMLCEngine("gemma-3-270m-it-q4f16_1-mlc", { appConfig });

Model Details

This is the Gemma 3 270m instruction-tuned model developed by Google converted to MLC format for usage in MLC WebLLM in the browser.

Model Description

Developed by: Google
Shared by [optional]: Aadeshisdoingsomething
Model type: Gemma

Model Sources [optional]

Repository: google/gemma-3-270m

Software

Converted for use in browsers using WebLLM. WebLLM documentation: https://webllm.mlc.ai/docs/user/get_started.html