Open in interactive viewer → charts, metric popovers & call review

Hesai's Second Act Arrives Before the First Act Grows Up

Revenue crossed the 50% non-auto threshold, but a Q3 miss and a soft Q4 guide show the Physical AI pivot is still a promise, not a parachute.
HSAI · Earnings Call · 2026-11-10

The Lidar Engine Sputtered

For three quarters, Hesai has been telling investors it is becoming something more than a lidar supplier. This quarter, it finally has the revenue mix to prove the point — and a guidance miss to complicate it. Total revenue hit RMB 1,068 million, up 34% year-over-year, but below the RMB 1.1-1.15 billion the company guided in August. Lidar shipments of roughly 730,000 also fell short of the 800,000-850,000 guide. The culprit was the automotive market: several customers cut 2026 sales targets, and ADAS volumes came in light.

We came in a little below previous expectations on ADAS, but third quarter total revenue still grew 34% year-over-year, while total lidar shipments increased 65%.

Peng Fan, CFO · 2026-11-10
That is a real crack in the core lidar story. The company still grew, still earned a GAAP profit of RMB 81 million — its sixth consecutive quarter — and still expanded core lidar segment operating income 88% to RMB 145 million. Gross margin was 41%. But the Q4 guide of RMB 1.1-1.15 billion assumes no snapback. CFO Andrew Fan: “For the fourth quarter of 2026, we expect total net revenues of RMB 1,100 million to RMB 1,150 million, reflecting more moderate downstream demand.” — Peng Fan, CFO · 2026-11-10 The growth algorithm is slowing, at least in autos.

The 50% Illusion

The headline number management wants you to focus on is the crossover: more than half of Q3 revenue came from outside automotive lidar. “businesses beyond automotive lidar contributed more than 50% of our total revenues this quarter” — Peng Fan, CFO · 2026-11-10. That is true, but it needs unpacking. The lidar segment still generated RMB 978 million, or 92% of total revenue. The non-auto portion includes robotics lidar — which sits inside that lidar segment — not just the new SGI businesses. The actual SGI segment, home to actuation modules and Kosmo, contributed RMB 91 million, or 8.5% of revenue. Still, SGI is growing fast: up 101% sequentially, YTD RMB 135 million, already above the RMB 100 million full-year target set in January. Management reaffirmed FY26 SGI revenue of RMB 200-300 million and a 2027 target of roughly USD 100 million with breakeven. That 2027 target has been raised twice — from RMB 500 million in May to USD 100 million/RMB 700 million in August — and is now being held. “we currently anticipate around RMB 100 million in full year 2026 revenues. Looking further out, we see a clear scaling path, potentially reaching around RMB 500 million in 2027.” — Peng Fan, CFO · 2026-05-19 The raise is real; the repeat is not new.

Actuation, Kosmo, and the Related-Party Question

The most tangible new disclosure is actuation. Hesai shipped over 23,000 modules in Q3, nearly doubling sequentially, with cumulative shipments above 35,000. Monthly capacity is 10,000 units, and management thinks annual shipments could reach 6 digits in 2027. A full humanoid can use more than 100 modules and account for over half of BOM cost. The anchor customer is Sharpa, a humanoid maker co-founded by CEO David Li. “Based on our current pipeline, we believe annual shipments could reach 6 digits in 2027.” — Yifan Li, CEO · 2026-11-10 The pitch is that if Hesai can meet humanoid requirements, it can sell into a much broader robotics market. But the early revenue is concentrated: Sharpa is the anchor customer and largest contributor. Management says it is in discussions with additional programs, but names are withheld. Kosmo, the 3D spatial data engine, is the other leg. It launched globally on October 14 and generated first revenue in Q3. The company has a partnership with Pixar's RenderMan and an initial program with Amap covering 1,000 sites. “Kosmo made its global debut in October 2026, secured deployments with world-class players such as Pixar's RenderMan, and has already started generating revenues.” — Yifan Li, CEO · 2026-11-10 The business model spans hardware, software subscriptions, and asset licensing. Management argues Kosmo could carry gross margins above 40% and eventually become recurring. That is plausible but unproven; current revenue is mostly hardware.

What Changed, What Didn't

The strategic pivot is not new. Physical AI was already the company's top keyword in Q1 2026; actuation and spatial intelligence dominated Q2 and Q3. What is new is the financial proof point: Kosmo revenue, 23,000 actuation units, and the 50% non-auto revenue mix. What also changed is the auto narrative. For years, Hesai's story was cost-down and penetration. Now management says the easy cost reductions are largely done. “the room for another major round of technology-driven cost reductions is becoming more limited. At this stage, cutting prices at any cost could come at the expense of product quality and reliability.” — Peng Fan, CFO · 2026-11-10 That is a mature-market statement. The bridge to growth is content per vehicle: more lidars per car, higher-end ET, overseas programs. The efficient ATX is being guided to stable pricing and margins in 2027. That is a reasonable bridge, but it depends on Level 3 adoption and overseas OEMs ramping. On the call, management cited design wins with 10 global OEMs through Chinese JVs, 7 Chinese OEM export programs, and the Mercedes-Benz L3 program. Those are real, but timing is back-end loaded. Meanwhile, the humanoid robot opportunity is enormous and early. The company's own 2027 SGI revenue target of USD 100 million is less than 10% of 2026 revenue, by our math. The phrase "100x larger than lidar alone" is a TAM statement, not a forecast. The balance sheet and profitability give Hesai room to invest. Six straight GAAP profitable quarters, core lidar cash generation, and SGI losses of RMB 83 million this quarter are manageable. But the valuation debate is now about what deserves a multiple: a profitable but slowing auto lidar leader, or a venture-stage robotics infrastructure platform. Q3 did not settle it. It gave both sides evidence. The stock's reaction will hinge on whether investors treat the Q3 miss as a one-off auto cycle or the start of a slower core. The second act is arriving; the first act is not dead. But for the first time, Hesai's non-auto revenue is big enough to matter and small enough to disappoint.