Space Exploration Technologies Corp. CEO Elon Musk stated on Thursday that the company could potentially be worth more than the entire planet Earth if it accomplishes its goals. The comment was made on social media platform X in response to a user discussing a deal between Anthropic and xAI, which involves Anthropic paying $45 billion through May 2029 to SpaceX for compute capacity at the Colossus and Colossus II data centers.

Musk countered the user's skepticism about the deal, emphasizing the long-term valuation potential of SpaceX. "You don’t seem to understand that SpaceX will be worth more than the rest of Earth if we accomplish our goals," Musk posted.

Starmind Constellation Details

SpaceX also revealed technical specifications for its new AI1 satellite, part of the Starmind constellation dedicated to AI compute. The satellites are designed to support chips from all providers and feature deployable liquid radiators to reject heat into the vacuum of space.

The AI1 satellite specifications include a wingspan of 70 meters and a 150 kW solar array. The constellation will be interconnected via laser links to beam AI compute back to Earth's surface through the Starlink satellite internet service.

AI1 Satellite Specifications

Musk has previously advocated for orbital AI compute as the only method to overcome Earth-based bottlenecks, stating, "Space is the only way to scale at scale."

SpaceX shares were up 0.06% at $152.25 during the after-hours session on Thursday.

SpaceXAI launched Grok 4.5 on Wednesday, introducing a low-cost artificial intelligence model designed to handle coding, complex work tasks, and research-heavy projects. The new model is priced at $2 per million input tokens and $6 per million output tokens, positioning it as a cheaper alternative to existing enterprise AI solutions. This pricing strategy aims to address the rising costs enterprises face with high-volume AI workloads.

SpaceXAI operates as a wholly owned artificial intelligence unit of Space Exploration Technologies Corp. The company stated that Grok 4.5 is its strongest model to date and was trained in collaboration with Cursor. The model can build applications from simple prompts, create Excel models, draft PowerPoint slides, and write documents in Word.

Neil Shah, an analyst at Counterpoint, noted that enterprises are experiencing "token bill shock" as autonomous agents and coding tools consume large volumes of tokens. Grok 4.5 enters the market as a fast, "good enough" option priced below Anthropic's Claude Opus 4.8, which costs $5 for input and $25 for output per million tokens.

Shah indicated that enterprises are moving toward diversified AI stacks, routing workloads based on cost, speed, and accuracy rather than relying on a single provider. Companies might use Claude for complex, high-stakes tasks while utilizing Grok for high-volume developer workflows and repetitive agentic routing.

Access to Cursor telemetry data could help Grok improve through developer interaction feedback. If Grok maintains its cost advantage while narrowing the accuracy gap, it could reshape enterprise AI economics and pose a pricing threat to OpenAI and Anthropic.