multimodel inputs for image does not work

[qkl9527]

dev env:
Google chrome canary 136.0.7089.0
Optimization Guide On Device Model: 2024.9.25.2033

const session = await ai.languageModel.create({
  expectedInputs: [
    { type: "image" }
  ]
});

# use HTMLImageElement
const userDrawnImage = document.querySelector(".testpng");

const response1 = await session.prompt([
  "explain this image:",
  { type: "image", content: userDrawnImage }
]);

console.log(response1);

response1 state is pending for a long time，No error reported

[qkl9527]

sometime report errror: Uncaught （in promise）AbortError: The request was cancelled.

[domenic]

This appears to be a bug report against a specific browser, instead of an issue with this explainer. Please open it on that browser's bug tracker instead.