← Back

Wire · market

SpaceXAI launches Grok 4.5 model for coding, agentic tasks

Published

10 July 2026

Topic

market

Sectors

AI Frontier ModelsAI Infrastructure

Geography

United States

Source

Read at reuters.com

Verified

Fusion42 · 10 July 2026 · Fusion42 review

SpaceX AI launches Grok 4.5, a coding and agentic task model trained on tens of thousands of Nvidia GB300 GPUs, priced at $2 per million input tokens and $6 per million output tokens—undercutting Anthropic's Claude Opus 4.8 on input cost whilst matching OpenAI GPT-5.6 Luna on output pricing. The launch follows SpaceX's $60 billion acquisition of Anysphere (Cursor) and consolidates SpaceX's vertical integration in enterprise AI tooling.

This Wire brief sits within Fusion42's coverage of AI Frontier Models and AI Infrastructure.

◆ The Wire takeaway

If you're building on top of Claude or GPT for coding tasks, SpaceX just made your foundation layer 60% cheaper on input. Cursor's now owned by SpaceX, trained on Grok 4.5, and available today—your customer switching cost just dropped to zero.

Coverage

1 source · 10 Jul 2026

Related on Wire

Topics

AI Frontier ModelsAI Infrastructureai-modelsgrok-4.5coding-agentspricing-warcursor-integrationvertical-integration