← Models

Grok 4.5

xAI · Grok

xAI's current flagship for coding, agentic knowledge work, and tool-using reasoning, with a 500K context window.

Part of Grok family · Other versions: Grok Build 0.1 , Grok 4.3 , Grok 4.20 , Grok 4.1 Fast , Grok Code Fast 1 , Grok 4
Type
multimodal
Context
500K tokens
Max Output
N/A
Status
current
Input
$2/1M tok
Output
$6/1M tok
API Access
Yes
License
proprietary
reasoning coding agentic tool-use multimodal structured-outputs long-context
Released July 2026 · Updated July 10, 2026

Overview

Freshness note: Model capabilities, limits, pricing, and regional availability can change quickly. This profile is a point-in-time snapshot last verified on July 10, 2026.

Grok 4.5 is xAI’s current flagship model for coding, agentic tasks, knowledge work, and technical reasoning. xAI launched it on July 8, 2026 and made it the new default model in Grok Build, while also exposing the grok-4.5 model through the xAI API.

The launch uses SpaceXAI as the public masthead and product branding after SpaceX acquired xAI in February 2026. The API domain, environment-variable names, legal terms, and current footer identity still use xAI or X.AI LLC, so SpaceXAI should be treated as the public brand around the combined organization rather than a clean replacement for every xAI technical identifier.

Capabilities

Grok 4.5 accepts text and images and produces text. xAI positions it for software engineering, long-horizon agent work, scientific and mathematical reasoning, and professional knowledge tasks. The model supports:

  • Configurable low, medium, and high reasoning effort, with high as the documented default.
  • Function calling and structured outputs for application-controlled agent loops.
  • xAI server-side web search, X search, and code execution tools.
  • Image-aware coding and analysis workflows alongside text-only repository work.
  • Responses API and Chat Completions API integrations.

The launch reports strong provider benchmark results and roughly 80 tokens per second in xAI’s serving environment. Those are useful launch signals, but teams should validate quality, latency, and tool reliability on their own repositories and workloads.

Technical Details

Current official model documentation lists:

  • Model ID: grok-4.5
  • Aliases: grok-4.5-latest and grok-build-latest
  • Context window: 500,000 tokens
  • Modalities: text and image input, text output
  • APIs: Responses and Chat Completions
  • Tools: function calling, structured outputs, web search, X search, and code execution

xAI does not publish a separate maximum output-token value on the model page at this snapshot. Here, maxOutput: 0 means “not separately published,” not a literal zero-token limit. Set explicit application-level output, timeout, and tool budgets.

Pricing & Access

Published API pricing per 1M tokens:

  • Input: $2.00
  • Cached input: $0.50
  • Output: $6.00

xAI notes that requests above 200K tokens use higher long-context pricing. Verify the current pricing page and region before committing a production budget.

Grok 4.5 launched across the xAI API console, Grok Build, Cursor, Microsoft Office add-ins, and selected gateway and cloud partners. xAI also announced limited-time free access through Grok Build and Cursor.

Regional availability is an important constraint: the July 8 launch says Grok 4.5 was not yet available in the EU through SpaceXAI products or the xAI API console, with an expected rollout later in July. EU teams should confirm live availability instead of assuming the global launch applies to their account.

Best Use Cases

Choose Grok 4.5 for:

  • Difficult coding-agent work that benefits from image input, tools, and deeper reasoning.
  • Multi-step research or knowledge workflows using xAI search and code execution.
  • Scientific, mathematical, and engineering analysis where an xAI-native flagship is useful.
  • Grok Build workflows where the current default model is preferred over a pinned older route.

Keep tasks bounded, review diffs, and test outputs before deployment. For cheaper xAI API coding calls, Grok Build 0.1 remains a separately priced route. For lower-cost general agent traffic, Grok 4.3 remains relevant.

Comparisons

  • Grok 4.3 (xAI): Lower-cost previous flagship with a larger 1M context window; 4.5 is the stronger current route for coding and agentic knowledge work.
  • Grok Build 0.1 (xAI): Cheaper 256K coding API model; Grok 4.5 is now the default inside the Grok Build CLI.
  • GPT-5.6 (OpenAI): Competing current flagship family with stronger OpenAI product integration and governance documentation.
  • Claude Sonnet 5 (Anthropic): Competing coding and agentic model with a mature Claude Code workflow and different safety and platform trade-offs.