<!-- mobian-agent-page publisher="dailydev" canonical="https://daily.dev/posts/browser-based-image-inpainting-runs-locally-if-one-doesn-t-mind-a-big-download-slr2ockwc" -->

---
title: Browser-Based Image Inpainting Runs Locally, If One...
description: Simon Willison ported the Moebius 0.2B image inpainting model to run entirely in a web browser using WebGPU, with no server-side GPU required. The project...
canonical: https://daily.dev/posts/browser-based-image-inpainting-runs-locally-if-one-doesn-t-mind-a-big-download-slr2ockwc
twitter:card: summary_large_image
twitter:site: @dailydotdev
og:type: website
og:site_name: daily.dev
og:title: Browser-Based Image Inpainting Runs Locally, If One Doesn’t Mind A Big Download | daily.dev
og:description: Simon Willison ported the Moebius 0.2B image inpainting model to run entirely in a web browser using WebGPU, with no server-side GPU required. The project...
og:url: https://daily.dev/posts/browser-based-image-inpainting-runs-locally-if-one-doesn-t-mind-a-big-download-slr2ockwc
og:image: https://api.daily.dev/og/posts/Slr2ockwc.png
og:image:alt: Browser-Based Image Inpainting Runs Locally, If One Doesn’t Mind A Big Download
og:image:width: 1200
og:image:height: 630
og:locale: en
---

> ## Documentation Index
> Fetch the complete documentation index at: https://daily.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Browser-Based Image Inpainting Runs Locally, If One Doesn’t Mind A Big Download

**[Hackaday](https://daily.dev/sources/hackaday)** · 2 min read · 1 upvotes · 0 comments

## Summary

Simon Willison ported the Moebius 0.2B image inpainting model to run entirely in a web browser using WebGPU, with no server-side GPU required. The project involved converting a PyTorch model to ONNX format and leveraging the CacheStorage API for a 1.27 GB model download. Notably, Claude Opus 4.8 (via Claude Code) handled the entire coding process autonomously, serving as a real-world test of how AI coding agents have matured. Willison reflects on what he learned through the process despite writing no code himself.

## Full article

daily.dev links to this article rather than hosting it. Read it at the original source: <https://hackaday.com/2026/07/10/browser-based-image-inpainting-runs-locally-if-one-doesnt-mind-a-big-download>

## Questions this post answers

### How can I run an image inpainting model locally in the browser without a server?

A PyTorch-based image inpainting model called Moebius 0.2B can be converted to ONNX format and run client-side using WebGPU acceleration in supported browsers. The model file, about 1.27 gigabytes, is downloaded once and cached using the browser's CacheStorage API so subsequent visits do not require re-downloading it.

_daily.dev surfaces experiments like this for developers exploring in-browser ML inference with WebGPU and ONNX._

### Can Claude Code build a full project autonomously without a developer writing any code?

Yes, in one documented case Claude Opus 4.8 running through Claude Code built an entire browser-based image inpainting tool autonomously, including converting a PyTorch model to ONNX, wiring up WebGPU inference, and handling large-file caching, without the developer writing any of the code themselves. The agent even generated its own document explaining the resulting architecture.

_developers gauging how far ai coding agents have come can follow real project write-ups on daily.dev._

## Similar posts on daily.dev

- [Porting the Moebius 0.2B image inpainting model to run in the browser with Claude Code](https://daily.dev/posts/porting-the-moebius-0-2b-image-inpainting-model-to-run-in-the-browser-with-claude-code-oaowh9jtk) · Simon Willison · 1 upvotes · 0 comments

---

Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#claude-code](https://daily.dev/tags/claude-code), [#webgpu](https://daily.dev/tags/webgpu)

[View this post on daily.dev](https://daily.dev/posts/browser-based-image-inpainting-runs-locally-if-one-doesn-t-mind-a-big-download-slr2ockwc)

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]}
{"@context":"https://schema.org","@type":"TechArticle","headline":"Browser-Based Image Inpainting Runs Locally, If One Doesn’t Mind A Big Download","url":"https://daily.dev/posts/browser-based-image-inpainting-runs-locally-if-one-doesn-t-mind-a-big-download-slr2ockwc","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/browser-based-image-inpainting-runs-locally-if-one-doesn-t-mind-a-big-download-slr2ockwc"},"datePublished":"2026-07-10T11:20:01.522Z","dateModified":"2026-09-13T19:14:18.092Z","description":"Simon Willison ported the Moebius 0.2B image inpainting model to run entirely in a web browser using WebGPU, with no server-side GPU required. The project...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/0213a6c12256bf895ef5b0add2187d54?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/0213a6c12256bf895ef5b0add2187d54?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Hackaday","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Hackaday","logo":"https://media.daily.dev/image/upload/s--JhDfhy70--/f_auto,q_auto/v1767541814/logos/hackaday","url":"https://daily.dev/sources/hackaday"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/browser-based-image-inpainting-runs-locally-if-one-doesn-t-mind-a-big-download-slr2ockwc","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai-agents,claude-code,webgpu","timeRequired":"PT2M"}
{"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Hackaday","item":"https://daily.dev/sources/hackaday"},{"@type":"ListItem","position":3,"name":"Browser-Based Image Inpainting Runs Locally, If One Doesn’t Mind A Big Download"}]}
{"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/browser-based-image-inpainting-runs-locally-if-one-doesn-t-mind-a-big-download-slr2ockwc#faq","mainEntity":[{"@type":"Question","name":"How can I run an image inpainting model locally in the browser without a server?","acceptedAnswer":{"@type":"Answer","text":"A PyTorch-based image inpainting model called Moebius 0.2B can be converted to ONNX format and run client-side using WebGPU acceleration in supported browsers. The model file, about 1.27 gigabytes, is downloaded once and cached using the browser's CacheStorage API so subsequent visits do not require re-downloading it. daily.dev surfaces experiments like this for developers exploring in-browser ML inference with WebGPU and ONNX."}},{"@type":"Question","name":"Can Claude Code build a full project autonomously without a developer writing any code?","acceptedAnswer":{"@type":"Answer","text":"Yes, in one documented case Claude Opus 4.8 running through Claude Code built an entire browser-based image inpainting tool autonomously, including converting a PyTorch model to ONNX, wiring up WebGPU inference, and handling large-file caching, without the developer writing any of the code themselves. The agent even generated its own document explaining the resulting architecture. developers gauging how far ai coding agents have come can follow real project write-ups on daily.dev."}}]}
```

