xAI prioritizes AI coding tools and its Grok Bot agent product
xAI has shifted focus to AI coding tools and its Grok Bot product, with Grok Build, Grok 4.7 and Grok Bot 1.0 now leading development priorities.
The company, which now operates under the rebranded name SpaceX Super Intelligence, has been building toward this developer-focused strategy throughout 2026. Their new priority list centers on a terminal coding agent, a model optimized for extended programming sessions, and a team-oriented bot designed for workplace chat environments.
Grok Build and Grok 4.7
The first component is Grok Build, which launched in beta in May 2026 as a terminal-based coding agent built for complex programming tasks. Originally limited to SuperGrok and X Premium+ subscribers, it is now powered by Grok 4.7, xAI's leading model for coding and agentic tasks that launched on September 21, 2026.
Grok 4.7 features a 500,000-token context window, allowing it to maintain a large portion of a codebase in view simultaneously. This capability supports long-running agentic workflows that span multiple files and steps. By November 10, the model had been experimentally integrated into AWS's Kiro platform, extending its presence across IDE, CLI and web environments.
Grok Bot 1.0 for Teams
The newest addition is Grok Bot 1.0, announced around November 11, which targets team environments rather than individual developers. It integrates with Slack and Microsoft Teams, enabling groups to share information through the bot and have it proactively handle tasks without waiting to be prompted.
The idea is that groups can share information through the bot and have it handle tasks proactively rather than waiting to be asked.
Bundled Strategy
xAI is bundling Grok, Grok Bot and access to Cursor into various X subscription tiers rather than selling them as standalone products. This strategy positions xAI both as a competitor and a repackager of the same market, as Cursor is named as a direct competitor alongside Anthropic's Claude Code.
Early models like Grok Build 0.1 were priced at as low as $1 per million input tokens and $2 per million output tokens. The bundling approach offers developers potentially lower costs but may create confusion about what they're actually paying for, especially if they only need specific tools.
Related
