OmniVision Group's Q3 2026 revenue reached 8.124 billion yuan, a record high, with net profit up more than 50% quarter-on-quarter
OmniVision Group's Q3 2026 revenue was 8.124 billion yuan, up 3.8% year-on-year, setting a new single-quarter record; net profit attributable to shareholders was 1.079 billion yuan, up 50.53% quarter-on-quarter, and gross margin recovered to 30%. Cumulative revenue for the first three quarters was 22.149 billion yuan, with semiconductor design business revenue of 16.346 billion yuan, accounting for 73.80%, while the share of intelligent driving, professional imaging, machine vision and robotics product lines continued to rise. The company expects Q4 revenue of 7.759 billion to 8.324 billion yuan, up 9.71% to 17.71% year-on-year.
Body · AI translation
On the evening of October 28, the 2026 third-quarter report disclosed by OmniVision Group showed that the company achieved operating revenue of 8.124 billion yuan in the third quarter, a year-on-year increase of 3.8%, setting a new historical high for the company's single-quarter revenue; net profit attributable to shareholders of the listed company was 1.079 billion yuan, a month-on-month surge of 50.53%; gross margin recovered to 30%, up 1.73 percentage points quarter-on-quarter.
Against the backdrop of periodic fluctuations in the semiconductor industry and pressure on end demand, the highlights of this report card are not limited to the numerical recovery, but more importantly to the clear main line behind it—product upgrades driving recovery, with the revenue structure continuing to improve.
Recovery curve: single-quarter revenue rises quarter by quarter, profitability clearly repaired
Looking over a longer period, OmniVision Group's operations since the beginning of this year have shown a typical quarter-by-quarter climbing trend: first-quarter revenue was 6.414 billion yuan, second-quarter 7.611 billion yuan, and third-quarter further rose to 8.124 billion yuan, with single-quarter revenue growing consecutively quarter-on-quarter and hitting a historical high. In the first three quarters, the company achieved cumulative revenue of 22.149 billion yuan, a year-on-year increase of 1.68%.
The repair on the profit side is even more evident: single-quarter net profit attributable to parent rose from 503 million yuan in the first quarter to 717 million yuan in the second quarter, and reached 1.079 billion yuan in the third quarter, a quarter-on-quarter increase of more than 50%; total net profit attributable to parent in the first three quarters reached 2.299 billion yuan. At the same time, single-quarter gross margin also recovered from 29.38% in the first quarter and 28.27% in the second quarter to 30% in the third quarter, basically comparable to the gross margin in the same period last year (30.34%).
It is worth mentioning that on the same day the third-quarter report was disclosed, the company also released a forecast announcement of main operating data for the fourth quarter: it is expected to achieve revenue of 7.759 billion yuan to 8.324 billion yuan in the fourth quarter, a year-on-year increase of 9.71% to 17.71%, with gross margin of approximately 29.44% to 30.74%. This means that the year-on-year growth rate of fourth-quarter revenue is expected to rise significantly.
Structural improvement: design business revenue structure optimized, high-end share continues to increase
Breaking down the revenue structure makes the quality of this financial report clearer.
In the first three quarters, the company's semiconductor design business achieved revenue of 16.346 billion yuan, accounting for 73.80% of operating revenue, and remains the absolute main business. More importantly, the structural changes within the design business itself: the revenue share of the intelligent driving, professional imaging, machine vision and robotics product lines continued to increase, and the share of high-end products expanded, directly driving the company's third-quarter gross margin to steadily recover quarter-on-quarter.
The confidence for structural optimization comes from sustained R&D investment. In the first three quarters, the company's R&D expenses were 2.229 billion yuan, a year-on-year increase of 5.91%, and the proportion of R&D expenses in revenue remained above 10%. These investments are accelerating into new products and new growth points along directions such as intelligent driving, professional imaging, machine vision and robotics, edge AI, medical, security, and optical communications and AI computing infrastructure.
Product matrix: intensive moves across four major battlegrounds, high-end new products entering the harvest period
If financial figures are the result, products are the true protagonist of this third-quarter report.
Since 2026, OmniVision Group has centered on its core capability of "visual perception" and intensively launched new products across four major battlegrounds: intelligent driving, machine vision and robotics, edge AI, and medical imaging, with many products already moving from launch to mass production.
In the intelligent driving field, high-end automotive CIS continues to lead. OX08D30, released in late September, is the latest flagship of the company's TheiaCel™ technology family. This 8-megapixel automotive image sensor combines LOFIC lateral overflow integration capacitor with DCG™ dual conversion gain to achieve LED flicker mitigation (LFM) under all lighting conditions, with a maximum detection distance of 200 meters, while power consumption is reduced by 20% and size by 50% compared with similar solutions.
With aCSP™ wafer-level packaging, OX08D30 is pin-compatible with the previously launched OX08D10, helping automakers smoothly upgrade perception hardware. It has now been sampled and entered the mass production stage. In terms of platform ecosystem, the company announced at the beginning of the year that OX08D10 and OX03H10 gained support from NVIDIA's DRIVE AGX Hyperion platform and were included in the Thor platform; OX08D10 had previously completed pre-integration with Qualcomm's Snapdragon Ride platform.
In addition, OAX7700, released by the company at the end of September, is an ultra-low-power AI coprocessor ASIC for Always-On Vision applications: the built-in NPU can perform human detection, object recognition and distance measurement locally, with overall power consumption reduced by 97% compared with similar solutions on the market, and supports 5 to 10 seconds of pre-recorded video buffering; the 5.5mm×3.5mm 60-pin BGA package is suitable for space-constrained automotive cameras and can upgrade existing automotive exterior camera solutions such as surround view systems (SVS) into parking monitoring systems based on ultra-low-power edge AI. The product is planned for mass production in the first quarter of 2027, reserving a new engine for the edge AI wave.
OmniVision Group, founded 31 years ago, is consolidating its intelligent driving base with a two-wheel drive of "high-end products + leading platforms"; the statements in the fourth-quarter operating data forecast announcement that "the automotive business continues to recover" and "automotive analog chips are being introduced into mass production" further confirm the improving trend of this curve.
In the machine vision and robotics market, building "eyes" for embodied intelligence. OG05D, released on the same day as the third-quarter report, is the industry's first 2.2μm pixel global shutter single-exposure HDR image sensor.
At 5 megapixels, the new product achieves a maximum frame rate of 146fps in linear mode, an increase of 143% over the previous generation; with DCG™ on-chip combination technology, a single exposure at 60fps can achieve HDR of more than 70dB with no motion artifacts; Nyxel® near-infrared technology enables quantum efficiency at 850nm to exceed 55% and quantum efficiency at 940nm to increase by more than 35%, greatly enhancing low-light and night-vision capabilities; the compact 6.6mm×5.3mm package makes it easy to embed in robotic arm ends and humanoid robot bodies.
From barcode scanning and factory automation to humanoid robots and robot training data collection, OG05D precisely positions itself in the two major tailwinds of smart manufacturing and embodied intelligence. At present, the company's OG01A, OG02D and OG05B are already in mass production, and three new products, OG0FA, OG01N and OG05D, have been sampled and are expected to enter mass production in the fourth quarter, forming a complete product echelon.
In the medical imaging track, a disposable endoscope solution has also taken shape. The OVMed™ ultra-thin cable imaging module with LED illumination released in June integrates wafer-level optics, illumination and cabling into one, and together with the OCHTA, OCHFA, OCHSA, OCH2B and other CameraCubeChip™ series, provides a "turnkey" solution for disposable endoscopes, and complies with IEC 60601 and ISO 13485 medical standards. At present, five CameraCubeChip products—OCHTA10, OCHFA10, OCHFA20, OCHSA10 and OCH2B10—have all been launched, and the company has formed complete solution capabilities from chips to modules in this high-growth track.
Competitive landscape and growth outlook
In terms of industry position, according to September data from third-party institutions such as TSR, OmniVision Group rose for the first time in 2025 to become the world's second-largest CIS (CMOS image sensor) supplier, with its automotive CIS market share firmly ranking first globally. Combined with its H-share listing in January this year, which raised HK$5.318 billion to supplement R&D and capacity reserves, the company's path of evolving from image sensors to a multi-wheel-driven platform of "vision + analog + optoelectronics" is becoming increasingly clear.
Looking ahead, the forecast year-on-year growth rate of fourth-quarter revenue guidance is further elevated compared with the third quarter, intelligent driving maintains relatively fast growth, and emerging applications such as professional imaging, machine vision and robotics, and edge AI continue to innovate. As new products such as OG05D gradually enter mass production in the fourth quarter and automotive analog chips and high-speed analog electrical chips for optical communications are introduced at scale, this visual perception leader known for product innovation is expected to continue delivering recovery and growth in the visual upgrade wave driven by AI.