Meta's Muse Wonder ships without a single cost-vs-score plot, researchers cry foul
AccBalanced · x · 2026-10-14
Meta launched Muse Wonder, a new family of frontier models claiming SOTA performance for agentic work, but the release conspicuously lacks any cost-vs-score plots in the blog or eval cards — only bar charts. "A very confusing way to release a model at best, and dishonest at worst."
The lone available Artificial Analysis scatter plot shows Muse Wonder (Max) scoring slightly below Gemini 4 (high) while costing more than 2x. Critics argue cost-performance curves should be standard for a "major" model release.
More from Models
- Gemini 4 Argon shows strong cyber capabilities: 87% ExploitBench, 68% CWE-Bench — scaling01 · 2026-10-14
- Early users wowed by Gemini 4 Argon's 3D design quality — prajdabre · 2026-10-14
- EmbeddingGemma2 on CPU matches OpenAI's embedding-3-large in 96-config retrieval benchmark — tomaarsen · 2026-10-14
- Physical Intelligence launches Pi API beta for fine-tuning π robotic foundation models — chelseabfinn · 2026-10-14
- Researcher says he had Argon attempt the Indus Valley script, with limited progress — adityakusupati · 2026-10-14
- Leaked Gemini 4 Argon benchmark shows trends matching Astra — scaling01 · 2026-10-14