Li Auto (LI) Q2 2026 earnings review
Margin Bleed Slows, But Growth Remains Elusive
Li Auto delivered a mixed Q2, showing sequential recovery from a disastrous Q1 but remaining far below its historical profitability. Total revenue fell 15.1% YoY to RMB 25.7 billion, generating a net loss of RMB 1.7 billion. The good news: the product transition is gaining traction, lifting vehicle margins to 9.4% from Q1's 6.1%. The bad news: despite aggressive launches of the new Li L6 and refreshed L8/L9 models, Q3 delivery guidance of 95,000-100,000 units implies virtually zero sequential volume growth. The shift from high-margin EREVs to a lower-priced portfolio has fundamentally altered the company's earnings power.
🐂 Bull Case
Gross margin and vehicle margin both expanded sequentially (to 11.0% and 9.4% respectively). The successful launch of higher-trim L9 models and stabilized production indicates the worst of the product-mix margin compression is likely over.
The successful rollout of the proprietary MACH M100 AI chip and MindVLA model solidifies Li Auto's technological moat, justifying premium pricing and driving future software monetization.
🐻 Bear Case
Despite a massive product refresh cycle, delivery guidance is flat. The new L6 and updated L-series appear to be cannibalizing existing sales rather than aggressively taking new market share.
Free cash flow remains negative (RMB -1.3 billion) while operating margins are deeply negative (-9.0%). The company is burning cash while heavily repurchasing shares.
⚖️ Verdict: 🔴
Bearish. While sequential margin improvement provides some relief, a company priced for growth guiding for flat QoQ deliveries and sustaining heavy net losses warrants significant caution.
Key Themes
Stagnant Volume Growth Contradicts Product Optimism
Management highlighted 'robust order flow' for the new Li L6 and the success of the L-series upgrade. However, the Q3 guidance of 95,000 to 100,000 deliveries implies a Stable, flat sequential growth profile compared to Q2's 98,330 units. This specific data point contradicts the narrative of a revitalized growth engine, suggesting the new models are primarily cannibalizing older ones rather than expanding the customer base.
Sequential Margin Recovery
Reversing the catastrophic Q1 collapse, vehicle margins improved from 6.1% to 9.4%. CFO Tie Li expects further margin expansion in H2 2026 as the product mix optimizes with higher sales contributions from the premium Livis trim and the all-new Li L9.
Intense Market Competition Weighing on Pricing
CEO Xiang Li cited 'intense market competition' in China's NEV market. This macro-level price war is evident in the financial results: while Q2 deliveries fell 11.5% YoY, vehicle sales revenue fell much harder (-16.7% YoY). This delta points to severe pricing pressure and a lower average selling price (ASP) required to move metal.
Livis Day and the Next-Gen AI Architecture
Li Auto is accelerating its transition from a pure hardware manufacturer to an 'embodied AI' company. At Livis Day, they launched the MACH Mind-Pro and MACH Mind-Edge language models, the MACH VLA machine intelligence model, and the proprietary MACH M100 dynamic dataflow AI chip. This tightly integrated tech stack is intended to reduce reliance on off-the-shelf suppliers like NVIDIA and drastically lower computing costs while boosting ADAS performance.
Li L6 Launch Targeting the Broad Market
The July rollout of the new Li L6 (priced aggressively at RMB 249,800) is intended to dominate the RMB 200,000-to-300,000 SUV segment. Equipped with an EREV-dedicated 51 kWh LFP battery and the MACH M100 chip, it provides a high-volume anchor for the brand to offset the slower sales of its high-priced BEV offerings.
Capital Allocation Paradox: Buybacks Amidst Cash Burn
Free cash flow was negative RMB 1.3 billion in Q2, representing the fourth consecutive quarter of operating cash flow struggling to clear heavy capex requirements. Despite this Decelerating liquidity profile, management aggressively repurchased US$150.9 million worth of shares in Q2. While they boast an RMB 87.5 billion cash fortress, burning cash on buybacks during a heavy, unprofitable product transition raises capital allocation questions.
Other KPIs
Decelerating loss sequentially from Q1's RMB 3.0 billion gap, but a massive reversal from the RMB 827.0 million operating profit generated in Q2 2025. Operating margin sits at -9.0%, underscoring the heavy toll the lower ASPs and transitioning product mix have taken on operational leverage.
Stable. R&D remained essentially flat YoY (-1.2%) and QoQ (+2.0%). This demonstrates management's commitment to fully funding the VLA large model, AI chips, and BEV architectures despite the steep drop in top-line revenue.
Guidance
Stable sequentially compared to the 98,330 vehicles delivered in Q2. On a YoY basis, it represents an Accelerating growth profile (+1.9% to +7.3%) purely because Q3 2025 was a notably weak base (93,211 deliveries). It implies the new product barrage is failing to trigger a volume breakout.
Stable. This projects a YoY change of -2.8% to +2.3%, and implies slight sequential growth versus Q2's RMB 25.7 billion. The fact that revenue is guided up slightly while deliveries are guided flat implies management expects ASPs to rise as the high-trim L9 models make up a larger portion of the sales mix.
Key Questions
Volume Cannibalization
With Q3 delivery guidance pointing to flat sequential growth despite the ramp-up of the highly anticipated L6 and the refreshed L8/L9, to what extent are these new models cannibalizing existing sales rather than capturing new market share?
Sustainable Margin Floor
Vehicle margins recovered to 9.4% in Q2, but still sit roughly 10 points below last year's average. With the lower-priced L6 expected to drive a massive portion of volume, what is the realistic, long-term gross margin ceiling for the company?
Capital Allocation Strategy
The company reported negative free cash flow again this quarter but spent over $150 million on share repurchases. Why is management prioritizing buybacks over preserving capital during an unprofitable hardware transition phase?