Baidu Stock Price Forecast: AI Infrastructure And Kunlunxin IPO Push Upside From $148 To $165

Baidu (NASDAQ:BIDU): AI Infrastructure, Chip IPO And Deep Value Behind A Noisy Earnings Trough

Current NASDAQ:BIDU Setup: $148 Stock Pricing A -7% Revenue Year Like A Stagnant Ad Platform

Baidu (NASDAQ:BIDU) trades around $148.07 after a pullback from a recent high near $151.08, versus a 52-week low at $74.71. That move has already doubled the equity from the lows, yet the market cap of roughly $51B still values the company as if it were a slow, ad-heavy Chinese internet name rather than a full AI stack player. The current trailing P/E of ~43.5 reflects a depressed earnings base distorted by a RMB 16.2B long-term asset write-down in Q3 2025, while forward numbers imply a much lower earnings multiple in the mid-teens. Anyone looking at the headline RMB 31.17B Q3 revenue, down 7.1% YoY, and a reported RMB -11.23B net loss with a -36.0% margin, sees an ugly quarter; but those figures are dominated by non-recurring accounting and a shrinking legacy ad engine, not by the AI infrastructure business that is actually driving the next leg of growth. The equity is effectively priced like a challenged search/ads franchise while the underlying economics increasingly come from cloud, proprietary chips, and AI infrastructure that carry a different multiple in global markets. For real-time pricing, the stock can be tracked on NASDAQ:BIDU live chart.

Revenue Mix In NASDAQ:BIDU: 75% Advertising Shrinks 18% While Baidu Core And Cloud Grow 21%

Baidu’s revenue profile still looks old-economy internet if you only glance at the mix: roughly three-quarters of total sales come from online marketing and search, and just one-quarter from non-advertising services. That is why a lot of global investors still label NASDAQ:BIDU as a “mature” Chinese web portal and assign it a compressed multiple. The dynamics are the opposite of what the headline mix suggests. Online marketing revenue in Q3 ran at about RMB 15.3B (≈$2.16B) and fell 18% YoY, confirming that the classic ad engine no longer carries the growth story. Baidu Core – which consolidates the non-ad AI businesses and much of the infrastructure stack – grew 21% YoY to roughly RMB 9.3B (≈$1.31B). Within that, Cloud AI revenue reached RMB 6.2B, also up 21% YoY, showing that the runway is now in infrastructure and AI services that are monetized, not in search pageviews. The market is still fixated on the shrinking 75% bucket instead of paying for the 21% growth in the non-ad 25% that is rapidly compounding and shifting the company’s economic center of gravity.

AI Cloud And Infrastructure Inside NASDAQ:BIDU: 33% Infra Growth And 128% Accelerator Subscriptions

The most important line inside NASDAQ:BIDU today is the AI infrastructure stack layered into Baidu Cloud. Within the RMB 6.2B cloud number, the infrastructure component – platforms and compute sold to enterprises and government – delivered around RMB 4.2B and expanded 33% YoY. That is not a toy project; it is a real infrastructure P&L at double-digit growth. Even more critical is the subscription layer on Baidu’s in-house accelerators. Subscription revenue linked to these AI accelerators surged 128% YoY, which shows a pivot from project-type contracts to recurring, usage-based revenue. This is the same direction the U.S. market rewarded in hyperscalers: recurring AI infrastructure, not one-off consulting. Baidu is now selling computing power, inference and real applications into both private and public sectors on commercial terms. For a stock that still trades like a discounted Chinese ad name, a 33% growth infrastructure segment with triple-digit subscription growth is mispriced. The company’s own vision reinforces that: management explicitly frames the stack as three blocks – cloud infrastructure, AI applications, and AI-driven marketing services – instead of product-by-product, which is exactly how scaled AI operators are being valued elsewhere.

Kunlunxin Spin-Off: NASDAQ:BIDU Unlocks A Domestic AI Chip Pure Play At A Private Value Of ~RMB 21B

The single biggest new catalyst for NASDAQ:BIDU is the announced IPO of its AI chip arm Kunlunxin in Hong Kong. Baidu owns about 59% of this unit, which designs AI accelerators used to run large language models, support Baidu’s ERNIE stack, and power AI workloads inside Baidu Cloud. In the last funding round in July 2025, Kunlunxin raised RMB 2B (≈$280M) at a valuation of around RMB 21B (≈$3.0B), up RMB 3B from the previous round. That private value is almost certainly below what a hot on-exchange AI chip listing can command in the current Chinese market, where peer IPOs such as Moore Threads posted first-day gains of more than 400% and other AI names like Biren also exploded out of the gate. Beijing’s industrial policy forces Chinese manufacturers to source at least 50% of their chip supplies domestically. That rule shifts structural demand toward local AI silicon providers like Kunlunxin and away from foreign GPU providers, turning Kunlunxin from a Baidu side-project into a policy-aligned national asset. With Kunlunxin’s chips already embedded into Baidu’s own infrastructure and large-scale AI deployments, the spin-off does not just create a number on a spreadsheet; it forces the market to assign an explicit multiple to AI hardware that was previously buried inside consolidated results of NASDAQ:BIDU.

Baidu’s AI Stack Around NASDAQ:BIDU: Full-Spectrum Capability From Chips To Cloud To ERNIE

Unlike many internet platforms that simply call external APIs, NASDAQ:BIDU operates a vertically integrated AI stack. At the base sits Kunlunxin’s AI accelerators, designed for training and inference workloads in large language models and other high-intensity AI tasks. On top of that Baidu Cloud provides compute, storage and model hosting, and then the application layer includes products such as the ERNIE chatbot and AI-driven marketing solutions. The same chips that Kunlunxin plans to bring to market through the IPO in 2026–2027 are already deployed in Baidu’s own infrastructure, meaning that commercial usage is verified, not theoretical. Cloud AI revenue of RMB 6.2B and the 128% surge in accelerator subscription income show that the stack is being monetized across multiple layers. For NASDAQ:BIDU, the strategic benefit is that the firm can compete as a true AI platform – chips plus cloud plus models plus apps – instead of being locked into foreign vendors’ roadmaps. That is precisely the architecture Beijing wants to see inside Tier-1 national champions as it tightens domestic content rules.

Apollo Go And Autonomous Miles: Operating Scale Without Being Priced Into NASDAQ:BIDU Yet

The robotaxi business Apollo Go is not driving the valuation for NASDAQ:BIDU today, but the operating metrics are well past the “demo” stage. Apollo Go has delivered more than 3 million fully autonomous rides and operates in 22 cities, with over 240 million autonomous miles logged. Those are scale statistics that, in any other market, would already invite separate capital allocation debates. The reason the equity market still treats Apollo Go as a “nice-to-have” is that unit economics and regulatory trajectory are not yet fully clear, so most fundamental models sensibly treat it as an out-of-the-money option. The important point for investors is not to assign zero; it is to recognize that NASDAQ:BIDU embeds an autonomous mobility asset with millions of rides and hundreds of millions of miles of data, on top of the AI Cloud and Kunlunxin stories, at a time when the quoted multiple still looks like it belongs to a cyclical ad business.

Q3 2025 P&L Behind NASDAQ:BIDU: One-Off RMB 16.2B Write-Down Masks Operating Transition

Looking at the reported RMB 31.17B Q3 revenue and the RMB -11.23B net loss, NASDAQ:BIDU appears to be in severe trouble. Net profit margin printed at around -36.0%, EPS fell to RMB 1.39, down an extreme 91.6% YoY, and EBITDA collapsed to roughly RMB 3.39B, a 56% drop. Taken at face value, those numbers would argue for a value trap. Part of that picture is real: the online marketing engine is shrinking, revenue is down 7.1%, and margins are under pressure because the company is spending heavily on cloud hardware, content and go-to-market. The distortion is the RMB 16.2B (≈$2.25B) long-term asset write-down booked in the quarter. Once that one-time charge is stripped out, the underlying profitability looks more like a compressed margin AI transition than an operational collapse. Costs of revenue increased 12% YoY, largely due to cloud and content, and SG&A also climbed 12%, with credit losses and sales spend contributing. Baidu Core’s non-GAAP EBITDA margin fell to around 9%, which is clearly lower than historical levels but consistent with a company pre-investing in capacity before utilization and leverage catch up. For a stock trading near $148, the question is whether investors accept lower near-term margin as “capex through the P&L” in AI, or insist on value-stock type free cash flow in the middle of a platform transition.

Balance Sheet And Liquidity: RMB 296.4B War Chest Underpinning NASDAQ:BIDU’s AI Capex Cycle

The investment case for NASDAQ:BIDU depends heavily on whether the company can fund its AI build-out without blowing up the balance sheet. On that front, the numbers are clear. Total assets sit around RMB 444.1B, up 6.4% YoY, against total liabilities of roughly RMB 156.5B, up 13.3% YoY. Equity stands at about RMB 287.6B, with cash and short-term investments alone at roughly RMB 124.8B, only 6% lower year-over-year. Total liquidity and investments reach approximately RMB 296.4B, giving Baidu a substantial buffer to absorb lower margins and elevated cloud hardware spend. On the cash-flow side, the Q3 snapshot shows RMB 1.26B cash from operations (down about 70.7% YoY), RMB 3.0B from investing (up 121.5%), and RMB 4.68B from financing (up 186.7%), resulting in a net cash increase around RMB 8.73B, up 155% YoY, even through a trough earnings quarter. Free cash flow was negative around RMB -1.34B, down about 139%, again reflecting the heavy AI investment phase. With a price-to-book ratio quoted near 0.19, the equity market effectively values NASDAQ:BIDU at a fraction of its book value despite this liquidity cushion – an anomaly for a company that is not fighting for survival but reallocating capital into high-growth AI infrastructure.

Sum-Of-The-Parts Logic For NASDAQ:BIDU: Ads Worth ~$65, New AI Businesses Another ~$35, Cash Roughly ~$50 Per ADS

A sum-of-the-parts framework is more appropriate for NASDAQ:BIDU than a single blended multiple, because the business is now three distinct economic engines. The legacy advertising and search segment, still about 75% of revenue but in structural decline, can be valued conservatively at 12–13x 2025 earnings, below Chinese online peer averages. On that basis, the ad/search operations are worth roughly $65 per ADS. The cloud and non-ad AI businesses – including Cloud AI, AI-based marketing services, and the wider infrastructure platform – are growing in the low-to-mid-20s percent, with recurring elements expanding faster. An EV/Sales multiple of about 5.5x, well under U.S. infrastructure peers, yields around $35 per ADS for this bucket. The third piece is liquidity and investments: with almost RMB 300B in cash and investments and a conservative holding company discount of 30%, that pool is worth roughly $50 per ADS. Adding these three components gives a base fair value close to $150 per ADS, very close to where NASDAQ:BIDU recently traded around $150–151. That number does not reflect options like Apollo Go or the chip IPO; it is effectively the floor valuation for the current configuration.

Upside Case For NASDAQ:BIDU: Lower Cash Discount And Slight Cloud Re-Rating Push Fair Value Toward $159–$165

A more realistic, still conservative, upside case for NASDAQ:BIDU comes from tweaking two inputs in that SOTP. First, applying a 20% discount to liquidity instead of 30% adds roughly $7 per ADS, recognizing that a blanket 30% haircut on a very liquid asset pool is excessive for a company of this scale. Second, a modest 0.2x uplift in the EV/Sales multiple for the cloud and AI infrastructure segment – reflecting higher quality recurring revenue as accelerator subscriptions keep compounding at triple-digit rates – contributes roughly $2 per ADS. Those adjustments move implied fair value from around $150 to roughly $159 per ADS. A separate valuation cross-check via a P/E framework leads to similar territory. At a forward P/E of 20x on consensus next-year EPS near $8.24, fair value lands around $165 per share, implying upside from the current $148 band. Both approaches converge on a value range of roughly $160–165 as a rational medium-term target if the AI execution stays on track and the chip IPO achieves anything close to peer enthusiasm.

Relative Valuation: NASDAQ:BIDU Trading At ~15x Forward Earnings Versus U.S. AI And Cloud Names At 25–30x

Even after an 82% share price increase in the last year, NASDAQ:BIDU still trades at a forward P/E of roughly 15x, which corresponds to an earnings yield of around 6.6%. Comparable Chinese large-cap tech peers such as Alibaba sit nearer 16.6x forward earnings (earnings yield about 6%), while U.S. AI infrastructure and cloud leaders often command 25–30x forward P/E or more. On an EV/EBITDA basis, Baidu changes hands in the high 7x range, whereas selected software/cloud benchmarks tracked on professional terminals trade in the mid-teens. The discount is partly justified – U.S. peers enjoy cleaner governance, deeper capital markets and higher margins – but the gap is wider than the risk delta. If NASDAQ:BIDU simply migrates from 15x to 20x forward earnings as the market starts to value AI cloud, Kunlunxin and recurring accelerator revenue more like U.S. infrastructure businesses, the stock price needs to move toward the $160–165 zone described earlier. At the current $148 quote, the market is paying China risk multiples for assets that look structurally closer to global AI infrastructure peers than to legacy portals.

Key Risks Around NASDAQ:BIDU: Ad Decay Speed, Cloud Profitability And Geopolitics

Owning NASDAQ:BIDU is not a free ride. The first execution risk is the slope of the legacy ad decline versus the ramp of the cloud and AI businesses. If online marketing revenue, already down 18% YoY, decays too quickly while capital expenditures and operating costs for AI infrastructure keep rising at the current pace, the company could face a prolonged period of negative free cash flow. The second risk is the quality of cloud growth: 21% revenue expansion and 33% infrastructure growth are attractive, but if margins stay thin and operating leverage fails to materialize, the market may continue to treat Baidu Cloud as a low-return experiment rather than a high-multiple asset. The third, and potentially dominant, risk factor is geopolitical. NASDAQ:BIDU is exposed to Chinese regulatory shifts, U.S. policy toward Chinese listings and semiconductor exports, and possible sanctions or blacklisting events that could move the stock independently of fundamentals. The recent easing of U.S. GPU export restrictions to China may also re-open the door for Nvidia hardware, increasing competition for Kunlunxin’s chips. Any disappointment on China’s domestic semiconductor content rules, delays in Kunlunxin IPO approval, or shocks to Baidu’s search franchise could compress the multiple again despite solid AI execution.

Investment View On NASDAQ:BIDU: AI Infrastructure-Driven Re-Rating Candidate, Rated Buy With $165 Target

Pulling the strands together, NASDAQ:BIDU is not a clean growth story and not a simple value stock; it is an AI infrastructure transition sitting on top of a shrinking, cash-generative ad base. Q3 2025 shows a RMB -11.23B headline loss, a -36% margin, and a 7.1% revenue decline, but beneath that the non-ad Baidu Core business grows 21%, Cloud AI reaches RMB 6.2B with 21% growth, infrastructure revenue jumps 33%, and accelerator subscriptions accelerate 128%, supported by a liquidity buffer of roughly RMB 296.4B. Kunlunxin’s planned Hong Kong listing, at a prior private valuation of about RMB 21B, gives the company a direct way to surface AI chip value, while Apollo Go and other options are still not priced in. With the stock around $148, a conservative SOTP supports about $150 per ADS, and realistic upside scenarios cluster in the $160–165 band as forward P/E expands toward 20x on expected $8+ EPS. Given the asset base, balance sheet and AI trajectory, the risk/reward is skewed positively. The stance on NASDAQ:BIDU is Buy, with a working 12-month target of $165, recognizing that execution on AI cloud, Kunlunxin’s IPO and geopolitical stability will dictate how quickly the market closes the valuation gap.

That's TradingNEWS