SpaceX stock is back in focus after JPMorgan said it is becoming “increasingly positive” on Grok, with analyst Doug Anmuth maintaining an Overweight rating and a $240 price target on SPCX. The increasingly bullish view follows improvements in Grok’s performance, lower model costs and the planned integration of AI coding platform Cursor, which JPMorgan believes could accelerate SpaceX’s push into the lucrative enterprise AI market. With SPCX trading around the mid-$130s before Tuesday’s move, the $240 target implies roughly 75%-80% upside — but investors are being asked to balance rapidly accelerating AI revenue against staggering capital expenditures and another major stock unlock approaching in September.
The call matters because SpaceX is no longer simply a rocket-and-satellite company. After combining with xAI, the business now spans Starlink connectivity, launch services, AI models, X, enormous terrestrial data centers and plans for orbital computing infrastructure. Second-quarter AI revenue surged roughly 250% year over year to $2.6 billion, while Starlink remained the financial engine with $4.3 billion of quarterly revenue. JPMorgan’s increasing confidence in Grok therefore raises a much bigger question for SpaceX stock: could AI eventually become valuable enough to justify the extraordinary amount of capital SpaceX is pouring into compute?
JPMorgan Keeps a $240 Target on SpaceX Stock
JPMorgan’s Doug Anmuth maintained an Overweight rating and $240 price target, with the firm becoming more optimistic about the commercial prospects for Grok. Anmuth highlighted signs that incorporating Cursor-related data into supplemental training has already improved model performance, while Grok 4.6 reportedly combines increasingly competitive intelligence with meaningfully lower costs than rival models. That combination matters because price-performance is one of the biggest variables determining which models enterprises deploy at scale. If a company can obtain comparable results while materially reducing inference costs, switching or adding workloads becomes much easier to justify.
JPMorgan also sees enterprise customers becoming increasingly important to Grok’s monetization. That is potentially far more valuable than relying primarily on consumer chatbot subscriptions because companies can generate large, recurring workloads across coding, research, customer support and software-development functions. Anmuth’s thesis suggests Grok’s investment case is beginning to shift from “can the model compete?” toward “how quickly can SpaceX monetize a competitive model?” That is a meaningful transition for SPCX shareholders because the company’s AI spending is already being measured in tens of billions of dollars.
Cursor Could Be the Missing Enterprise Piece
Cursor sits at the center of JPMorgan’s enthusiasm. Seeking Alpha reported that the AI coding company had approximately $4 billion of annual recurring revenue as of June 2026, with around 75% coming from businesses, immediately giving SpaceX deeper exposure to enterprise customers rather than simply adding another consumer-facing AI product. SpaceX’s IPO prospectus originally disclosed an agreement allowing the company to provide Cursor with GPU compute while Cursor contributed personnel, datasets, workflows, prompts and technical know-how to improve Grok and potentially develop additional models collaboratively. SpaceX also disclosed an option structure valuing Cursor at an implied $60 billion equity value.
Related articles
The strategic logic is powerful. Grok supplies the foundation model, SpaceX supplies enormous computing capacity, and Cursor gives the combined ecosystem an established application used directly by software developers and enterprise engineering teams. Instead of trying to sell companies raw AI tokens and hoping they build useful products themselves, SpaceX could increasingly control multiple layers of the stack — infrastructure, models and applications. That vertical integration resembles the strategy that has made Nvidia so difficult to challenge in AI hardware, although SpaceX is attempting it across an even broader combination of compute and software.
Grok 4.6 Is Giving JPMorgan More Confidence
The analyst upgrade in sentiment is also tied directly to improving model quality. JPMorgan said it has seen tangible gains after Cursor data was incorporated into Grok training and believes Grok 4.6 delivers strong capabilities at significantly lower costs than some competing models. Lower cost does not guarantee adoption, and benchmark performance can vary substantially by workload, but economics are becoming increasingly important as corporations move AI applications from experimentation into production. A model that costs substantially less per completed task can produce major savings when usage reaches millions or billions of requests.
SpaceX itself has repeatedly emphasized rapid model iteration. Its public-offering prospectus said the company intends to continue scaling Grok through future generations including Grok 5, with plans for multi-trillion-parameter models and deeper reasoning capabilities. On the company’s August earnings call, management said Grok 4.5 had already produced strong enterprise feedback and that token consumption tripled after its July launch, while Grok 5 was expected before year-end. Those claims are management expectations rather than guaranteed outcomes, but they show how aggressively SpaceX is attempting to shorten AI development cycles.
The AI Business Is Already Becoming Financially Material
The strongest support for JPMorgan’s thesis comes from SpaceX’s actual financial results. Second-quarter revenue nearly doubled to $7.8 billion from $4.1 billion a year earlier, with AI revenue rising about 250%. Starlink accounted for more than half of total revenue, while AI became the fastest-growing major business inside the company. Total operating losses narrowed to $143 million from $970 million a year earlier, suggesting that rapid revenue growth is beginning to absorb some of the enormous fixed costs associated with SpaceX’s expansion.
S&P Global calculated that SpaceX’s AI segment generated approximately $2.6 billion of Q2 revenue, up 247.5% year over year and 213% sequentially, beating consensus expectations by roughly 27%. That is important because investors no longer need to value SpaceX’s AI business entirely on future promises. Compute contracts, Grok subscriptions, X-related monetization and other AI services are already contributing billions in quarterly revenue. The debate is now about whether revenue can eventually grow fast enough to justify the infrastructure being built to support it.
The Huge Problem: SpaceX Is Spending at an Almost Unthinkable Rate
The biggest threat to the bullish SpaceX stock forecast is capital intensity. SpaceX spent more than $18 billion on capital expenditures during Q2, including approximately $15.83 billion on AI infrastructure alone, compared with just $749 million of AI capex in the year-earlier period. CFO Bret Johnsen indicated capital spending could remain near those levels over the following couple of quarters. In other words, SpaceX spent roughly six times its quarterly AI revenue on AI infrastructure during the period.
Management argues that the economics will improve rapidly because deployed compute can start producing revenue almost immediately. SpaceX has told investors it expects less than a one-year payback period on new AI compute deployments and aims to have more than two gigawatts of capacity operating by the end of 2026, potentially approaching 10 gigawatts by the end of 2027. Those targets could create extraordinary earnings leverage if demand remains strong, but they also expose shareholders to significant execution and financing risk if utilization falls or AI pricing becomes more competitive.
Nvidia Is Critical to the SpaceX AI Strategy
SpaceX’s AI infrastructure strategy is also becoming tightly linked to Nvidia. Management said it plans to use Nvidia hardware exclusively for its expanding AI data centers, putting the company directly in the middle of the same supply-and-capex cycle driving hyperscaler spending at Microsoft, Meta, Alphabet and Amazon. This relationship gives SpaceX access to leading accelerator technology but also means enormous amounts of capital must be committed before Grok and outside compute customers generate corresponding revenue.
The scale of the wider market helps explain why SpaceX is willing to take that risk. JPMorgan estimates hyperscaler capital expenditures alone could reach roughly $697 billion in 2026, with AI infrastructure financing emerging as one of the defining capital-markets themes of the year. SpaceX is effectively betting that demand for AI compute remains constrained enough that it can profitably deploy almost every GPU cluster it builds. The company’s existing compute agreements with Anthropic, Google and other customers support that thesis, but a downturn in AI infrastructure demand would make the financial model far more vulnerable.
Starlink Still Provides the Financial Foundation
For all the attention surrounding Grok, Starlink remains the most dependable part of the SpaceX investment case. Connectivity revenue increased 66% year over year to approximately $4.3 billion in the second quarter, while subscriber numbers doubled to 12 million. Starlink’s operating income rose 79%, giving SpaceX a growing cash-generating business capable of supporting investment in AI and Starship. That diversification is important because SpaceX does not need Grok to finance the entire organization immediately.
The risk is that management keeps expanding investment faster than Starlink can generate cash. SpaceX has ambitions across AI, next-generation satellites, mobile connectivity, Starship, orbital data centers and potentially chip manufacturing. Each opportunity could be enormous individually, but pursuing all of them simultaneously creates a level of capital intensity rarely seen in public markets. JPMorgan’s Grok optimism therefore strengthens one piece of the valuation case without eliminating the fundamental question surrounding the entire company: when does enormous infrastructure spending turn into sustainable free cash flow?
A September Share Unlock Could Pressure SPCX Stock
There is also a near-term technical risk that has little to do with Grok’s performance. JPMorgan flagged that nearly 370 million SpaceX shares could become eligible for sale around September 9-10, potentially increasing the public float by approximately 20%. Lockup expirations do not automatically produce selling because employees and early investors can simply continue holding, but a sudden increase in available shares can pressure a recently listed stock if supply exceeds new investor demand.
That makes the timing of JPMorgan’s $240 target particularly interesting. The firm sees substantial fundamental upside while simultaneously warning that SPCX could face near-term volatility as additional shares become tradable. TipRanks currently shows a broader analyst consensus of 24 Buy ratings, five Holds and three Sells, with an average price target around $232.35. The bullish Wall Street consensus therefore remains intact, but investors chasing the Grok story may need to tolerate another potentially volatile stretch.
Outlook: Grok Is Becoming Much More Important to SpaceX Stock
JPMorgan’s increasingly positive stance on Grok matters because it reinforces a major transformation in the SpaceX stock investment thesis. SpaceX went public primarily known for reusable rockets and Starlink, but its fastest-growing financial engine is increasingly AI. Grok is improving, Cursor brings enterprise customers and software-development distribution, AI segment revenue has already reached billions of dollars per quarter, and SpaceX is constructing one of the world’s largest pools of compute infrastructure.
The opportunity is enormous, but so is the bill. Investors should watch Grok 5 performance, Cursor integration, enterprise adoption, AI segment revenue growth, utilization of new compute capacity and whether management’s promised sub-one-year AI infrastructure payback begins showing up in cash flow. The September stock unlock is another immediate risk. JPMorgan’s $240 target suggests Wall Street believes the current price leaves room for significant upside if those pieces fall into place.
SpaceX built its reputation by making rockets reusable and satellite internet global. If JPMorgan is right about Grok, the next chapter of SPCX may depend on whether Musk can make AI compute just as commercially powerful.
