Grok 4.5 Overview

Name: Grok 4.5
Brand: Grok 4.5

Grok 4.5 is a Large Language Models (LLMs) tool. Advanced AI model for coding, agentic tasks, and knowledge work with fast speeds. Key features include Advanced Coding Capabilities, Token Efficiency, and Fast Processing Speed. Best for software developers and engineers, data scientists and analysts and financial advisors and analysts.

⬆ 6 upvotes6 key features6+ alternatives →

About Grok 4.5

Grok 4.5 is SpaceXAI's most capable AI model built for coding, software engineering, and knowledge work. Trained alongside Cursor, it offers competitive pricing at $2 per million input tokens and delivers 80 tokens per second with a 500K context window.

Key Features

Advanced Coding Capabilities.

Built specifically for software engineering tasks including Rust, C/C++, and end-to-end app development from prompt to production with minimal specification.

Token Efficiency.

Uses 4.2 times fewer tokens than comparable models like Opus 4.8 on complex tasks, completing work in half the steps while maintaining quality.

Fast Processing Speed.

Delivers results at 80 tokens per second with a 500,000 token context window, allowing processing of entire codebases and long documents.

Competitive Pricing.

Priced at $2 per million input tokens and $6 per million output tokens, significantly cheaper than rival models while offering similar capabilities.

Office Integration.

Works with Word, PowerPoint, and Excel through Grok Build to create complex spreadsheets, diagrams, presentations, and documents with web research capabilities.

Multimodal Support.

Handles text and image inputs with function calling, structured JSON outputs, extended reasoning, and vision capabilities for analyzing screenshots and diagrams.

Frequently Asked Questions

Grok 4.5 is designed for coding, software engineering, agentic workflows, and knowledge work. It works well for building apps, debugging code, creating spreadsheets, writing documents, and handling complex technical tasks that require reasoning across large codebases.

Grok 4.5 costs $2 per million input tokens and $6 per million output tokens. Cached input tokens cost just $0.50 per million. This pricing is significantly lower than competing models like Anthropic's Opus 4.8 which costs $5 and $25 respectively.

Grok 4.5 performs competitively with models like Claude Opus and GPT-5.5 on coding benchmarks, though it may trail on some tasks. Its main advantage is cost efficiency, using about half the tokens per task while delivering results faster at 80 tokens per second.

You can access Grok 4.5 through the SpaceXAI API, Grok Build, Cursor on all plans, and the xAI console. It's available on desktop, web, iOS, CLI, and SDK. EU availability is expected in mid-July 2026.