Summary
Riley Brown breaks down a week in which Google and Meta both released new frontier models. On Monday, Meta launched Muse Wonder and Muse Wonder Full, which Meta’s benchmarks place just behind Claude Opus 5.5 on legal work, financial analysis, and long-horizon coding. Brown finds the pricing disappointing at $4 per million input tokens and $20 per million output tokens, the same as Opus 5.5, and notes it runs through a terminal tool rather than a polished coding app. Google followed on Tuesday with Gemini 4 Argon in the Gemini API, the Gemini app, and Antigravity, where access costs $100 a month on the Ultra plan.
To compare the models, Brown has them build a Call of Duty-style game, and the Gemini result is noticeably weaker than what Opus 5.1 produced earlier. He concludes that Claude Opus 5.5 still leads, even though many observers say Google and Meta have caught up with OpenAI and Anthropic.
The video also covers four Grok Bot updates, new Claude Code features, and a string of OpenAI Codex changes, including usage-limit resets, a larger context window for GPT 6.1 Sol rising from 272K to 500K tokens, and a roughly 60% smaller disk footprint for the desktop app. Brown closes with speculation about Anthropic’s rumored Fable 5.5.
📺 Source: Riley Brown · Published October 16, 2026
🏷️ Format: News Analysis
