xAI launched Grok 4.20 with a team of AI agents

xAI launched Grok 4.20, and it packs in a team of agents that work together on every request. Users watch them reason step by step through a new live thinking interface. The whole system works on the Grok website, including non-premium subscribers.

When you ask something, the agents kick in together. Grok leads the way while Harper, Benjamin, and Lucas each grab a piece of the job that matches their focus. They split the problem, think in parallel, and trade ideas back and forth.

The live thinking interface shows it all on screen. You see progress bars tick, quick notes pop up, and each agent builds on what the others say. Timers count down as they process at the same time. One pulls fresh details, another checks the logic, and they combine everything into one strong answer. Complex questions turn clear because you follow the flow instead of waiting for a finished reply. It feels like a quick team huddle you can actually watch.

Grok 4.20 (beta) launched with multi-agent reasoning (Source – xAI)

xAI made sure free accounts unlock these agents without extra steps. You sign in and start using them immediately. Paid plans give faster replies or longer sessions, but the core agent experience and live view stay open to everyone.

In recent live trading challenges, Grok 4.20 versions posted steady gains of 10 to 12 percent while many rivals lost ground. They read shifting conditions fast and used real-time data to stay ahead. That performance shows how the system handles practical tasks beyond basic chats.

The update builds straight from earlier agent tools and keeps xAI focused on tight real-time ties. Conversations now feel dynamic because you stay connected to the process the whole time.

In summary, Grok 4.20 makes agents a visible, collaborative team available to free users. Go to grok.x.ai or open the chat on X, log in with your account, and try a tough question today. Watch the live thinking unfold and see how it changes the way you work.

Author’s take

xAI has already implemented multi-agent models with the Grok 4 lineup, which eases the research and in-depth analysis. That multi-agent model was limited to high-priced subscription but now it’s available for all of the users, including the free ones. So that every user can benefit from its powerful reasoning capabilities.