AI/TLDR

Grok Build 0.1

xAI's fast agentic coding model — 256K context, always-on reasoning, $1/$2 per 1M tokens.

Overview

Grok Build 0.1 (model id grok-build-0.1) is xAI's fast, purpose-built coding model for agentic software engineering. It is the current flagship of the Grok Code line, succeeding Grok Code Fast 1, and launched in public beta on the xAI API on May 20, 2026.

The model is the engine behind the Grok Build CLI — xAI's terminal coding agent that runs as an interactive TUI, headlessly in CI scripts, or through the Agent Client Protocol (ACP) inside other apps. grok-build-0.1 is tuned to act as an autonomous engineering agent: it invokes tools, edits multi-file codebases, supports Model Context Protocol (MCP) servers, and runs in continuous loops until a task is done.

Grok Build 0.1 has a 256K-token context window and accepts both text and image input (text output only), so it can read diagrams, UI mockups, and error screenshots. Reasoning is always on — every response includes structured analysis before the final answer, and the reasoning effort is not user-configurable.

Released2026-05-20
LicenseProprietary
WeightsAPI only
Context256K
Max outputNo stated limit
ArchitectureReasoning model with always-on chain-of-thought; reasoning effort is not configurable. Powers the Grok Build CLI agent.
Knowledge cutoffNot disclosed
ModalitiesText, Vision
StatusAvailable (public beta)

Pricing

Input$1.00 / 1M tokens per 1M tokens
Cached input$0.20 / 1M tokens per 1M tokens
Output$2.00 / 1M tokens per 1M tokens

xAI API public-beta pricing for grok-build-0.1.

Pricing source ↗

Strengths

  • Trained specifically for agentic coding — tool use, function calling, and multi-file edits rather than chat
  • 256K context window holds an entire mid-sized codebase in memory
  • Always-on reasoning gives structured analysis before each answer
  • Text + image input — can read diagrams, mockups, and error screenshots
  • Low, flat pricing ($1 input / $2 output per 1M tokens) with $0.20 cached-input reads
  • Powers the Grok Build CLI with plan mode, headless/CI mode, MCP support, and ACP integration

Best for

  • Agentic coding workflows: refactoring, debugging, and feature work across large repositories
  • Driving the Grok Build CLI terminal agent interactively or headlessly in CI
  • Tool-calling and MCP-connected agents that plug into internal APIs and knowledge bases
  • Web development and multi-step build tasks that need a long context window
  • Reviewing visual inputs (diagrams, mockups, screenshots) alongside source code

How to access

ProviderModel ID
xAI ↗grok-build-0.1
OpenRouter ↗x-ai/grok-build-0.1

Grok Code — every version

The full lineage of the Grok Code line, newest first. Every version has its own page — click any to compare specs, benchmarks and pricing.

VersionReleasedContextLicense
Grok Build 0.1current2026-04Proprietary
Grok Code Fast 12025-08-28Proprietary

FAQ

What is Grok Build 0.1?

Grok Build 0.1 (model id grok-build-0.1) is xAI's fast, purpose-built coding model for agentic software engineering. It is the current model in the Grok Code line and powers the Grok Build CLI. It launched in public beta on the xAI API on May 20, 2026.

How much does Grok Build 0.1 cost?

On the xAI API it is priced at $1.00 per million input tokens and $2.00 per million output tokens, with cached input reads at $0.20 per million tokens.

What context window and modalities does Grok Build 0.1 support?

It has a 256K-token context window and accepts text and image input with text output. xAI's docs state there is no fixed output limit.

Does Grok Build 0.1 use reasoning?

Yes. Reasoning is always on and is not configurable — every response includes structured analysis before the final output, which makes it well suited to multi-step agentic coding tasks.