XUNFEIHEALTH (02506) has delivered a robust first-half performance that outpaced market expectations, reporting a 49.4% year-on-year increase in revenue to RMB 446 million. The company, recognized as China's first listed medical large model specialist, also posted a 53.5% rise in gross profit to RMB 236 million, achieving a consolidated gross margin of 52.9%. Sales collections reached RMB 348 million in the period, up 49% year-on-year, with the net loss attributable to shareholders narrowing by 20%, signaling an improvement in overall operational quality.
As the 2026 "15th Five-Year Plan" kicks off, the medical AI sector is transitioning from pilot validation to large-scale deployment. The company is leveraging this favorable policy environment by channeling resources toward its annuity-style ARR revenue streams, marking a strategic shift from one-off project deliveries to a platform-based operational model. This approach supports the three core business segments, creating a synergistic structure characterized by a stable foundation, a leading growth engine, and improving profitability.
Where to Begin: The Core Business Landscape
The AI Diagnosis Assistant segment, serving as the company's fundamental pillar, generated RMB 202 million in revenue during the first half, a 47.5% increase with a gross margin of 56.1%. This momentum follows a national demonstration event in Hefei in July 2026, where the company's Intelligent Medical Assistant and regional imaging platforms received official endorsement for accelerated nationwide rollout. The company's AI-assisted diagnosis tool now covers 828 districts across 31 provinces, serving over 77,000 primary healthcare institutions, with a market share exceeding 80% in deployed regions. Having provided more than 1.2 billion AI assistance suggestions and generated 490 million standardized electronic medical records, the segment has solidified its dominance. Additionally, over 700 grade-A hospitals are utilizing the next-generation assistant, which has achieved a 91% adoption rate for automatically generated medical records and is now exploring usage-based recurring revenue models.
The AI Health Assistant segment emerged as the primary growth catalyst, with revenue surging 74.9% year-on-year to RMB 182 million, the fastest growth among all three segments. The medical imaging cloud business was a standout, with revenue increasing by over 200%, as its Anhui platform now connects more than 1,980 institutions and houses over 160 million imaging records, contributing to an estimated RMB 2 billion in annual medical insurance savings. The company is strategically expanding its footprint into broader diagnostic imaging services, including laboratory tests, pathology, and electrocardiograms, a move supported by recent national pricing policies that incentivize data standardization. Furthermore, the post-discharge patient management service is now operating routinely at leading hospitals such as Qilu Hospital of Shandong University and the First Affiliated Hospital of USTC, with strong paid user conversion rates and sustained ARR growth momentum.
Why the Focus on Recurring Revenue Streams?
The AI Digital Infrastructure segment contributed RMB 62 million in revenue, up 7.9%, with its gross margin improving significantly by 22.7 percentage points year-on-year. This segment has established deployments in major cities including Shanghai, Shenzhen, and Hefei, with its "AI-driven resident health profiling" project in Shanghai's Jing'an district being recognized in a World Health Assembly case library. The segment also supports disease control and regional collaboration platforms across multiple provinces. Underpinning these operations is the proprietary Harness dual-mode architecture, which ensures deterministic control and compliance without modifying model weights. This engineering foundation has enabled the release of the Spark Medical Large Model V3.5, trained on domestic computing power, which has achieved a 92.4% average improvement in six core capabilities and reduced the medical hallucination rate from 3% to 1.9%. The model topped the MedBench evaluation and ranked first in IDC's 2026 China Medical Large Model Technology Assessment.
The company's strategy is now firmly centered on transitioning to an operating model with recurring revenue. This transition is driven by two primary engines: the imaging cloud business, which benefits from national policies and fee-for-service mechanisms, and post-discharge patient management, which has validated a clear patient-payment logic across multiple hospitals. The accelerating data flywheel effect, fueled by 1.71 million daily AI feedback instances from doctors and 520,000 from patients, continuously strengthens the company's competitive moat. By adhering to rigorous medical standards and maintaining a long-term perspective, XUNFEIHEALTH is not only optimizing its own revenue structure but also establishing a replicable commercialization benchmark for the entire medical AI industry. Looking ahead, the company aims to solidify its position as a global leader in medical large models, playing an increasingly vital role in the intelligent upgrade of China's healthcare system.