Image: gHacks Tech News / ghacks.net SpaceXAI and Cursor Launch Grok 4.5, Their First Joint AI Model
Weeks after the $60 billion acquisition closed, SpaceXAI and Cursor shipped Grok 4.5 — a model built for software engineering, legal work, and finance that's now available across all Cursor plans.
SpaceXAI and Cursor released Grok 4.5 on July 8, the first AI model the two companies built together since SpaceX formally agreed to acquire Cursor in a $60 billion all-stock deal.
The model is not a general-purpose chatbot. SpaceXAI positioned it explicitly for software engineering, legal analysis, and financial work — the kind of multi-step, document-heavy tasks where context length and reasoning depth matter more than chat latency. According to the company, Grok 4.5 “can understand large amounts of information, reason through complex problems and complete multi-step tasks with little human input.”
Elon Musk called it “roughly comparable to Opus 4.7, but much faster,” and said engineers at Tesla and SpaceX are using it in production. The model was trained across tens of thousands of NVIDIA GB300 GPUs using large-scale stability techniques developed during the Colossus 1 buildout.
What it can do
For developers, Grok 4.5 is designed to operate across large codebases — generating production-ready code, debugging, and planning development tasks end to end. That puts it in direct competition with Claude’s Sonnet and Opus lines, as well as OpenAI’s Codex.
On the legal and finance side, SpaceXAI claims the model handles contract review, case file summarization, regulatory analysis, financial statement examination, due diligence, and market intelligence gathering. That’s a different wedge than most coding-first AI tools have tried.
Availability and pricing
Grok 4.5 is available now in Cursor on all paid plans, with elevated usage limits during the launch period. It’s also accessible through Grok Build and the SpaceXAI console directly.
SpaceXAI priced the standard version at $2 per million input tokens and $6 per million output tokens. A faster premium version costs $4 per million input tokens and $18 per million output tokens. For comparison, Claude Opus 4.8 runs $10 per million input and $50 per million output — so Grok 4.5 is substantially cheaper if performance holds up on real workloads.
One caveat: Grok 4.5 is not yet available in the EU.
The bigger picture
This is the first concrete product signal from the SpaceX-Cursor acquisition. The deal made waves when it was announced — SpaceX paying $60 billion for a two-year-old coding tool startup was either a bet on AI-native developer workflows or a play for talent and distribution. Grok 4.5 suggests both: it’s a model that wouldn’t exist without SpaceXAI’s compute, but it ships through Cursor’s installed base rather than starting from zero.
Cursor’s annual revenue had reportedly crossed $2 billion as of mid-2026, with roughly 25% of enterprise AI clients subscribing in some capacity. That’s a real distribution channel for a new model to land in.
Sources: