News
Research Report
The AI Transformation 100: Concrete Ideas for Fixing How We Work
Key Takeaways
- New coding model. Grok Build 0.1 released in xAI API public beta.
- Key capabilities. Supports agentic coding, tool calling and fast function execution.
- Developer impact. Offers cost-effective, high-speed coding support for AI-driven workflows.
SpaceX released Grok Build 0.1, marking the company's first dedicated AI coding model. It is available in public beta through the xAI API and powers the Grok Build command-line interface.
According to xAI Grok Build 0.1 was trained specifically for agentic coding tasks, including web development, debugging and Model Context Protocol (MCP) support. The company said the model generates at more than 100 tokens per second and is priced at $1 per million input tokens and $2 per million output tokens.
Beyond coding, xAI claims the model is a fast, economical option for general-purpose agentic and tool-calling use cases. It integrates with harnesses such as Grok Build, Cursor, Hermes Agent, OpenClaw, Kilo Code and OpenCode, and is also accessible through OpenRouter and Vercel AI Gateway.
Table of Contents
What Can Grok Build 0.1 Do?
Key capabilities xAI claims for Grok Build 0.1:
| Feature | How It Works |
|---|---|
| Agentic coding | Trained for web development, debugging and MCP support |
| Function calling | Connects the model to external tools and systems |
| Reasoning | Processes logic before generating a response |
| Speed | Generates at more than 100 tokens per second |
| Context window | Supports up to 256,000 tokens per request |
The Developer Workflow Becomes More Autonomous
AI coding tools now support multi-agent workflows that span the full software development lifecycle from design through deployment with minimal human intervention.
Assistants vs. Agents
| Traditional Assistants | Agentic Coding Systems |
|---|---|
| Suggest code snippets | Execute multi-step development tasks |
| Reactive autocomplete | Goal-oriented workflow execution |
| Short interaction windows | Long-session contextual continuity |
| Code generation only | Debugging, testing & refactoring support |
Developers at Anthropic, OpenAI and Google already use these systems to generate code, debug applications, explain legacy systems and automate software development workflows. In fact, Google recently claimed that over 75% of new code created inside its company is generated by AI.
Architecture & Integration
The Model Context Protocol standardizes how AI applications connect to external tools and data sources, including CLI access. Anthropic created MCP to address integration complexity that limits enterprise AI deployment scope.
Apple's integration of the Claude agent SDK into Xcode 26.3 shows how autonomous coding agents are being embedded directly into developer toolchains.
Recent SpaceX News
SpaceX is approaching what could be the largest IPO in history, having filed an S-1 on May 20, 2026, seeking more than $75 billion at a roughly $1.75 trillion pre-money valuation. A Nasdaq listing under "SPCX" is expected as early as June 12.
The filing reported $18.7 billion in 2025 revenue but a $4.9 billion net loss, with Morningstar noting that Starlink's $1.19 billion Q1 2026 operating profit is effectively subsidizing xAI's $2.47 billion quarterly operating loss. The company merged with xAI in February 2026 at a combined valuation of $1.25 trillion.
On the AI front, SpaceX has moved to monetize its Colossus supercomputer infrastructure. In May 2026, Anthropic secured access to more than 300 megawatts of capacity — over 220,000 NVIDIA GPUs — at Colossus 1, with interest in partnering on multiple gigawatts of orbital data center capacity.
In April 2026, SpaceX also struck a deal with AI coding startup Cursor, offering an option to acquire the company for $60 billion or a partnership-only arrangement for $10 billion. The agreement gave Cursor access to Colossus's 200,000 H100/H200 GPUs while filling a gap left by xAI's slower-than-expected progress on coding tools.