Meta Platforms (NASDAQ: META) shares surged 11.4% on Monday, marking their largest one-day percentage gain since April 2025 and taking the stock to its highest closing level since October 2025.
The shares closed at $741.24 after reaching an intraday high of $753, putting Meta firmly above the $700 level that investors had been watching as a key test for whether its recent recovery could develop into a more durable trend.
It closed at $665.75 on Friday.
The move was driven in large part by the early success of Meta’s newly launched personal AI assistant, Muse, which climbed to the top of Apple’s US App Store.
But the stock’s technical breakout also matters because investors had been looking for evidence that Meta could sustain its recovery after a volatile year.
Earlier this month, Dan Russo, co-chief investment officer at Potomac Fund Management, which owns Meta shares, said in a Bloomberg report:
“Eliminating a legal overhang and then having a positive product in a hot narrative is certainly enough to get the early stages of a trend change going,” adding that in order for Meta’s most recent ascent to be considered durable, the shares need to see a “sustained break above $700.”
Monday’s close provided that move above the threshold, although investors will now be watching whether Meta can hold the level following such a sharp one-day rally.
The Muse craze has sent Meta toward its best monthly performance in more than 13 years — up 29.5% since the beginning of the month and more than 40% above its 52-week low in March, when the company’s increased spending projections triggered a sharp selloff.
The stock was about flat during premarket trading on Tuesday, falling by 0.2%.
Meta now has “a story to tell” around AI
Meta’s AI spending has become one of the biggest issues hanging over the stock.
The company is on track to spend between $130 billion and $145 billion on capital expenditure this year, largely covering physical infrastructure such as data centers.
While its traditional advertising business continues to grow rapidly, investors had been uncertain about whether Meta could develop meaningful revenue streams from AI beyond improving its existing recommendation and advertising systems.
Muse is beginning to provide an answer.
Introduced on September 8, the personal AI agent is designed to execute multistep tasks, including finding information across accounts, filling out web forms and negotiating and making online purchases.
Users can interact with Muse through its dedicated app or WhatsApp.
Meta CEO Mark Zuckerberg has described the company’s broader AI ambition as a quest for “personal superintelligence” — technology that understands users deeply and can be highly customised.
Muse offers an early example of what that strategy could look like in practice.
Wells Fargo analyst Ken Gawrelski who raised his PT on Meta on Monday to $796 from $640, citing early traction of Muse, pointed to Sensor Tower data showing that Muse recorded a record 264,000 US downloads on September 19 and 448,000 daily active users on September 18.
Gawrelski said Meta’s recent AI model launches, including Muse Spark 1.3 and the Muse assistant, mean the company “now has a story to tell” around its AI capabilities and services.
How Muse gives Meta a new monetisation route
The app is free at its basic level, while additional subscription tiers cost $20 and $100 a month depending on usage.
That gives Meta a potential AI revenue stream that is not directly dependent on advertising.
Another opportunity is Meta Model API, through which developers and enterprise customers can pay for access to Meta’s underlying AI models.
Mizuho analyst Lloyd Walmsley said the release represents a meaningful step toward demonstrating positive financial returns on Meta’s substantial AI capital expenditures.
The early adoption numbers have strengthened that argument.
Muse overtook ChatGPT as the leading free iOS app in the US on Friday and recorded 730,000 downloads in roughly five days following its September 8 launch, according to Sensor Tower.
By Monday, the market intelligence company said Muse had surpassed 2.5 million downloads.
The app was also ahead of Anthropic’s Claude, xAI’s Grok and Meta’s existing Meta AI app on Apple’s US App Store.
How analysts are rerating Meta stock based on Muse
The market had re-rated Meta downward based on spending anxiety and legal risk, and those concerns were being overweighted relative to the underlying advertising business, which was growing rapidly and generating substantial cash.
Muse is changing that.
Wells Fargo revised its earnings-per-share estimates to $31.86 for 2027 and $38.75 for 2028, compared with previous estimates of $32.07 and $38.73.
The new $796 target is based on 25 times the firm’s 2027 earnings estimate, up from a previous multiple of 20 times.
That is broadly aligned with Meta’s current P/E ratio of 25.3.
A significant number of analysts and investors had argued that Meta was structurally undervalued relative to its peers in the Magnificent Seven.
Morgan Stanley’s Brian Nowak has compared Meta’s current setup with Alphabet’s position in late 2025, when the Google parent benefited from a favourable antitrust ruling followed by a series of new AI products.
Meta is in position to keep releasing new products, “which in aggregate could add $10+ to EPS,” Nowak wrote in an Aug. 30 note.
“We don’t believe these are priced in Meta right now,” he added, pointing to a valuation that remains below peak 2025 levels, when Meta was viewed as an “AI winner” and GPU investments were driving faster growth.
JPMorgan also upgraded Meta to overweight earlier this month, citing the “strong early traction” of Muse.
Analyst Doug Anmuth said the initial success shows why “Meta is well-positioned to deliver consumer-driven AI products to its base of about 4B users.”
“There’s still meaningful upside potential as Meta is in the early stages of releasing frontier models and AI-driven products beyond advertising,” Anmuth added.
Muse may have a moat even without the best model
The early success of Muse has also raised a more fundamental question about the AI competitive landscape.
Stratechery analyst Ben Thompson said Meta’s Muse could pose a bigger strategic threat to frontier AI companies such as OpenAI and Anthropic than previously estimated, Stocktwits said.
In his latest column, Thompson called Muse “by a significant margin, the best and most approachable personal agent product I have tried.”
His observation was particularly notable because Muse Spark 1.3 is not considered the state-of-the-art AI model.
“That is the bearish signal,” Thompson wrote, arguing that AI models may already be capable enough for companies to build compelling consumer products without leading the model race.
Thompson believes personal agents could develop stronger moats than conventional chatbots because users can feed them personal information and integrate them into everyday routines.
“Model capability is good enough that compelling products — products that actually have moats — can now be built,” he wrote.
That could have significant implications for Meta because the company does not necessarily need to dominate the underlying-model race to build a valuable AI business.
Its existing social platforms, massive user base and consumer distribution could become equally important competitive assets.
Earlier this month, Oppenheimer questioned whether Muse will have a meaningful financial impact in the near term.
Analyst Jason Helfstein maintained a Perform rating on Meta and said the firm does not view Muse as a “game changer.”
He questioned whether enough consumers will ultimately pay for the service, particularly as users may already be paying for competing products such as Google’s Gemini and OpenAI’s ChatGPT.
However, while chatbots are interchangeable, an AI agent that has been fed a user’s financial accounts, calendar, shopping history, email, and personal preferences is not.
The switching cost of moving from one agent to another grows with every piece of information the user shares and every task the agent completes successfully.
Meta, with approximately 4 billion users across its platforms and unparalleled knowledge of their social graphs and behavioural patterns, is better positioned to build that kind of personal context than almost any other company.
The company can introduce AI through Facebook, Instagram, WhatsApp and its other platforms rather than relying entirely on users to discover a standalone product.
Amazon blockade exposes the agentic AI challenge
The Muse story, however, also comes with an important complication.
Over the weekend, Amazon.com (AMZN) blocked Muse from shopping on its website, citing a violation of its terms of use and security concerns.
The dispute highlights a major obstacle for AI agents: they may be capable of performing tasks across the internet, but the companies operating those platforms can restrict their access.
Pivotal Research Group analyst Jeff Wlodarczak wrote that “We believe this pressure is likely to broaden as agentic AI increasingly replaces human search, comparison and routine decision making.”
Wlodarczak added that widespread use of AI agents could hurt digital advertising businesses, potentially encouraging companies such as Amazon to develop their own first-party agents.
For Meta, that creates a tension between Muse’s promise and its practical limitations.
An agent becomes more valuable as it can access more services, but those same services may have economic incentives to prevent outside AI agents from controlling customer interactions.
Meta’s earnings estimates offer a near-term test
The rally above $700 is also coming ahead of Meta’s next quarterly earnings report to be released next month, where Wall Street’s expectations could provide an important test of whether the stock’s recent momentum can continue.
According to Zacks, Meta is expected to report quarterly earnings of $6.39 per share, down 11.9% from the year-ago quarter.
Revenue is expected to reach $63.18 billion, representing year-over-year growth of 23.3%.
The divergence between revenue and earnings expectations is notable.
Meta’s advertising business continues to support strong top-line growth, but despite the optimism around Muse, the company’s heavy investment in artificial intelligence and related infrastructure is still putting pressure on profitability.
More importantly, the consensus earnings estimate has been moving in the wrong direction.
Zacks said the consensus EPS estimate for the quarter has been revised 3.4% lower over the past 30 days.
That creates a potential complication for META stock after its sharp move above $700.
Empirical research has historically found a strong relationship between changes in earnings estimates and near-term stock-price movements.
In other words, a rising share price accompanied by falling earnings expectations can create a disconnect that investors eventually have to reconcile.
The recent rally has been driven largely by the improving narrative around Muse, Meta’s broader AI product pipeline and the prospect of new monetisation opportunities.
But the company will ultimately need to demonstrate that those investments are translating into financial results.
“So, make sure to keep an eye on META going forward to see if this recent jump can turn into more strength down the road,” Zacks said.
Meta Connect becomes the next stock catalyst
The timing of the Muse rally puts Meta Connect firmly in focus.
The conference on Wednesday is expected to provide investors with further updates on Meta’s AI products and could offer additional evidence about the company’s product roadmap.
For META stock, the setup has therefore changed considerably in a matter of weeks.
Muse has given investors the first substantial product-level evidence. Meta Connect will be the next test.
The post Meta stock crosses $700 mark on Muse hype: can the rally last? appeared first on Invezz