Summary
xAI has released Grok 4.7, and creator Alex Finn argues it’s the first model purpose-built to run inside an agent harness rather than just a chatbot. The video breaks down what’s different about this release, including benchmark comparisons against Anthropic’s Opus 5, pricing that stays flat compared to Grok 4.6, and why Finn believes AI companies are shifting focus from standalone models toward agent products like Grokbot, OpenAI’s Muse, and similar tools.
The bulk of the video is a hands-on walkthrough of Finn’s personal Grokbot setup, where he organizes multiple AI personas — a project manager, developer, and designer bot — for each major area of his life and work. He shows how these bots can be grouped into team-style chats for status updates, and explains why he considers this multi-persona structure more powerful for daily ‘power user’ tasks than single-thread tools like ChatGPT or Claude.
Viewers interested in agentic AI workflows will come away with a concrete template for structuring their own bot teams inside Grokbot, along with Finn’s take on how Grok 4.7 stacks up against competing frontier models on cost and capability.”
📺 Source: Alex Finn · Published September 22, 2026
🏷️ Format: Tutorial Demo
