Meta’s new ‘Muse Wonder’ model: good for AI agents, bad for design

<p>I put Meta’s new Muse Wonder (and Wonder Full) frontier models through the How I AI bench, which covers nearly 20 tasks across prototyping, agentic workflows, coding, writing, and creative work. I scored everything with an LLM-as-a-judge pass and my own voice notes, then pulled together the first-look results you see here.</p><p><br></p><p><strong>What you’ll learn:</strong></p><ol><li>​Why Meta’s “personal and professional agent” framing mostly holds up, and where it doesn’t</li><li>​What active focus mode actually does on long-running tasks, and whether it’s real</li><li>​The three agentic tasks where Muse ranked first, and why those categories matter</li><li>​What Muse’s design output actually looks like (it’s not good)</li><li>​How it compares to Sonnet on voice assistant tasks</li><li>​Where Muse lands on SVG and short-form video versus Opus 5.5, Astra, and Fable</li><li>​Whether I’m actually putting Muse into any of my agent workflows</li></ol><p>—</p><p><strong>Brought to you by:</strong></p><p><a href="https://getdx.com/howiai" target="_blank" rel="ugc noopener noreferrer"><strong>DX</strong></a>—Engineering intelligence for the AI era</p><p><strong></strong><a href="https://atlassian.com/howiai"><strong>Jira Product Discovery</strong></a>—Prioritize with insights, build with confidence</p><p>—</p><p><strong>In this episode, we cover:</strong></p><p>(00:00) Intro and what Muse Wonder is</p><p>(01:19) Meta&#39;s frontier capability claims</p><p>(04:36) Muse code</p><p>(06:06) Benchmark overview and how the How I AI bench works</p><p>(07:51) Overall results and TL;DR verdict</p><p>(11:53) Agentic tasks</p><p>(13:36) Email writing and voice assistant performance</p><p>(14:35) Where Muse falls short</p><p>(18:28) SVG illustration results</p><p>(19:33) Short-form video editing</p><p>(21:18) Fast on one-shot work, slow on agent tasks</p><p>(22:15) Final verdict and where I&#39;ll use Muse Wonder next</p><p>—</p><p><strong>Tools referenced:</strong></p><ul><li>​ Meta Muse Wonder / Muse Wonder Full: <a href="https://research.meta.ai/blog/muse-wonder" target="_blank" rel="ugc noopener noreferrer">https://research.meta.ai/blog/muse-wonder</a></li><li>​ Muse Code: <a href="https://developer.meta.com/ai/lp/muse-code" target="_blank" rel="ugc noopener noreferrer">https://developer.meta.com/ai/lp/muse-code</a></li></ul><p>—</p><p><strong>Where to find Claire Vo:</strong></p><p>ChatPRD: <a href="https://www.chatprd.ai/" target="_blank" rel="ugc noopener norefe