SpaceX Q2 Revenue Hits $7.8B as AI Capex Accelerates

SpaceX posted 92% revenue growth in Q2, driven by AI cloud deals and Starlink expansion, despite heavy capital spending.
Key points
- SpaceX Q2 revenue reached $7.8 billion, up 92% year-over-year, with AI revenue tripling to $2.6 billion.
- Starlink generated $3.3 billion in Q1 sales and $1.2 billion in operating income, offsetting $2.5 billion AI losses.
- AI capital expenditures hit $7.7 billion in Q1, exceeding space and broadband capex combined, highlighting infrastructure costs.
Space Exploration Technologies reported second-quarter revenue of $7.8 billion, a 92% year-over-year increase, following its June IPO. The company’s stock has experienced significant volatility since listing, ranging from $105 to an all-time high near $226 before settling around $155. This price action reflects market uncertainty regarding the sustainability of the company's rapid growth and heavy capital expenditure.
The business model is split into three segments: connectivity, space, and artificial intelligence. While Starlink remains the primary profit driver, the AI segment is now the focal point for future growth. Investors are closely watching capital spending, as AI infrastructure costs have surged ahead of contracted revenue in previous quarters, creating a margin pressure that the company must manage as it scales.
Starlink anchors profitability amid AI losses
Starlink generated $3.3 billion in sales and approximately $1.2 billion in operating income during the first quarter, serving as the cash engine for the broader organization. In contrast, the space division, which includes launch services and Starship development, recorded a $662 million operating loss due to $930 million in research and development costs. The launch segment itself brought in $619 million in revenue, indicating that the heavy spending is focused on next-generation technology rather than current commercial launch operations.
The AI segment, which includes Grok and cloud hosting services, reported $818 million in first-quarter revenue but incurred a $2.5 billion operating loss. This divergence highlights the capital-intensive nature of the AI business, where infrastructure buildout costs significantly outpace early revenue generation. The company’s ability to convert this investment into profitable recurring revenue is the central risk and opportunity for shareholders.
AI cloud deals drive revenue surge
In the second quarter, AI revenue nearly tripled to $2.6 billion, driven by new cloud capacity agreements. Key customers include Anthropic, paying $1.25 billion per month, and Google Cloud, contributing $920 million per month. Additionally, a six-month contract worth $6.7 billion, rumored to involve the Department of Defense, and a $1.11 billion monthly deal with an unnamed customer starting in December, underscore the scale of SpaceX’s compute infrastructure business.
Management, including CEO Elon Musk and CFO Bret Johnsen, projects an annual recurring revenue run rate of $100 billion by year-end. Deutsche Bank analyst Edison Yu described this target as likely achievable, citing the integration of Cursor acquisition services with Grok and the expansion of cloud hosting deals. This aggressive growth trajectory relies heavily on securing long-term contracts that offset the high cost of maintaining GPU clusters and data center infrastructure.
Capital spending defines margin outlook
Capital expenditures for the AI segment reached $7.7 billion in the first quarter, significantly higher than the $1.05 billion spent on the space business and $1.3 billion on broadband. As reported by The Motley Fool, this capex is the critical metric to monitor ahead of the third-quarter earnings report. If infrastructure costs continue to outpace contracted revenue, margins will remain under pressure despite top-line growth.
Recent reports indicate SpaceX is exploring the purchase of customer and operational data from defunct AI startups. This strategy would allow the company to acquire training data and hardware at a discount, potentially improving the efficiency of its AI models while reducing the premium paid for new chips. Such moves aim to optimize the cost structure of the AI division, which is currently the largest driver of operating losses.






