Meta's superintelligence lab launches Muse Wonder, a frontier model tuned for personal agents
alexandr_wang · x · 2026-10-13
Alexandr Wang announces Muse Wonder, saying the research process starts with evals: an internal Personal Superintelligence Index (PSI) measuring what makes models useful for personal agents rather than external benchmarks alone. The shared benchmark table shows Muse Wonder Full leading on PSI (59.1), Harvey LAB-AA legal agent work (13.1), SWE-Marathon (57.5), and MemoryBench (85.5) versus Claude Fable 5.1, GPT-6 Astra, and Claude Opus 5.5. Highlights include Active Memory (best memory across days/weeks/months), Active Focus for task completion, and claimed SOTA in finance and legal domains.
Related event: Meta Unveils Muse Wonder, Its Most Capable Agentic Frontier Model(88 posts)→
More from coding & agent
- Muse Wonder scores ~42% on senior engineer benchmark vs Sol 6.1's 71% — danshipper · 2026-10-13
- Gemini CLI PR fixes sandbox: trap-wrapped commands broke per-command permission rules — MauroDruwel · 2026-10-13
- Alexandr Wang admits Muse Wonder wasn't trained on 3D game demos, insists it excels at real coding — alexandr_wang · 2026-10-13
- Sentry CEO on reverse engineering vs copyright: selling reimplementation invites lawsuits; LLM clean-room arguments fail — zeeg · 2026-10-13
- Dev to test $13 voice board with Codex and GPT Live in DIY AI device — OpenAIDevs · 2026-10-13
- Dev Ditches MacBook Pro: Runs opencode Coding Agent on Transparent-Screen Laptop — KevinNaughtonJr · 2026-10-13