APAC AI HealthTech: 8 Companies Compared

APAC is generating some of the most significant AI healthcare transactions in the world. Hims and Hers acquired Australian digital health company Eucalyptus for up to $1.15 billion in February 2026 — the largest digital health exit in Australian history. Eli Lilly signed a $1.3 billion licensing deal with Korean RNA-editing company Rznomics. NVIDIA acquired Vietnam’s VinBrain. XtalPi struck a $6 billion collaboration with DoveTree. Insilico Medicine IPO’d on the Hong Kong Stock Exchange raising $293 million, 1,427 times oversubscribed.

The APAC AI healthtech market is not emerging — it has emerged. This comparison maps eight AI-native healthtech companies across the region: what their AI actually does, where the M&A deals are happening, and what acquirers are paying.

---

The Scoreboard

Company	Country	Founded	Funding / Status	Sub-vertical
Insilico Medicine	HK / UAE	2014	IPO’d HKEX Dec 2025; $293M raised	AI drug discovery
XtalPi Holdings	China	2015	IPO’d HKEX Jun 2024; $400M pre-IPO	AI drug discovery + robotics
Lunit	South Korea	2013	KOSDAQ listed; $829M market cap	AI medical imaging / oncology
Harrison.ai	Australia	2018	$250M+ raised; Series C Jan 2025	AI radiology / multi-finding
Mirxes	Singapore	2014	SGX IPO May 2025	AI genomics / liquid biopsy
Doctor Anywhere	Singapore	2017	$195M raised	AI telehealth / care navigation
Deepwise Healthcare	China	2017	$115M raised	AI radiology / cancer detection
Nalagenetics	Singapore	2016	$13.6M raised (A*STAR spin-off)	AI pharmacogenomics

---

AI Differentiation Tier

---

Company Profiles

Insilico Medicine

Insilico is the most internationally visible APAC AI drug discovery company. Its Pharma.AI platform comprises three modules: PandaOmics identifies biological targets using generative models; Chemistry42 generates novel molecular candidates; inClinico predicts clinical trial outcomes. The company has been the first to take a fully AI-generated small molecule — rentosertib, an IPF treatment — through Phase 2a clinical trials.

The Hong Kong IPO in December 2025 raised HKD 2.28 billion (approximately $293 million) and was oversubscribed 1,427 times. The company reported $56.24 million in FY2025 revenue. Post-listing market cap reached approximately $2.7 billion, implying roughly 48x revenue — a premium that reflects the novelty of its generative AI drug platform and the scarcity of comparable companies at clinical stage.

Thirty programs across fibrosis, oncology, CNS, and cardiometabolic disease are in the pipeline. Ten are at clinical stage. The platform value significantly exceeds the pipeline value — a strategic acquirer is buying a generative drug discovery engine with a live pipeline option attached.

Strategic acquirer thesis: Global pharma (AstraZeneca, Roche, Pfizer) seeking to absorb a generative AI discovery engine, or APAC-focused pharma (Takeda, Sino Biopharmaceutical) looking to add AI-generated pipeline depth. HK listing provides a clean acquisition vehicle under HK Takeovers Code.

---

XtalPi Holdings

XtalPi combines quantum physics, AI, and automated robotic labs for crystal structure prediction and drug candidate generation. Its competitive differentiation from pure neural network approaches is the physical chemistry modeling — it uses quantum mechanics, not only pattern matching, to predict molecular behavior.

The company IPO’d on the HKEX in June 2024 and announced a landmark collaboration with DoveTree in 2025: $51 million upfront, $49 million deferred, and up to $5.89 billion in milestone payments — the largest AI drug discovery deal announced in APAC. H1 2025 revenue of RMB 517 million (approximately $71 million) represented 404% year-on-year growth, with the company reaching first-ever half-year profitability.

Partners include Eli Lilly and Janssen. The robotics lab automation angle creates a specific acquisition thesis for CROs (WuXi AppTec, Lonza) that want AI-powered chemistry integrated into their manufacturing infrastructure.

Strategic acquirer thesis: Global pharma wanting AI-robotics discovery integration, or a large CRO seeking to vertically integrate drug design capability. The milestone structure of the DoveTree deal actually makes XtalPi more, not less, attractive to acquirers — it proves the market will pay large numbers for this capability.

---

Lunit

Lunit produces two AI product lines: INSIGHT for radiology (chest X-ray, mammography, breast cancer risk prediction) and SCOPE for digital pathology (tumor microenvironment analysis in oncology). The company is KOSDAQ-listed with a market cap of approximately $829 million and TTM revenue through Q3 2025 of $54 million — a 3.7x year-on-year revenue increase.

The partnership architecture tells the acquisition story: AstraZeneca uses Lunit AI pathology for non-small cell lung cancer; Daiichi Sankyo has integrated SCOPE into two oncology drug development programs; Roche Diagnostics distributes SCOPE through its digital pathology ecosystem; Microsoft co-developed an AI platform launching in Q2 2026. Lunit is embedded in the workflows of companies that would be its natural acquirers.

An FDA 510(k) submission for Lunit INSIGHT Risk (five-year breast cancer prediction) was filed in late 2025, with clearance expected in 2026. US regulatory clearance will materially expand the buyer universe.

Strategic acquirer thesis: Roche (already a distribution partner), Philips, or Siemens Healthineers would be natural strategic buyers. A pharma company wanting to lock in AI pathology for companion diagnostics is an alternative. KOSDAQ listing creates mandatory tender offer obligations above 33% stake — price accordingly.

---

Harrison.ai

Harrison.ai’s Annalise Enterprise simultaneously identifies 124+ clinical findings on a single chest X-ray. The multi-finding approach — detecting pneumonia, pneumothorax, fractures, aortic abnormalities, and dozens of other conditions in one pass — is orders of magnitude more comprehensive than single-indication AI diagnostic tools. This makes Annalise a platform acquisition rather than a point solution.

The January 2025 Series C raised AU$179 million (approximately $112 million) with an unusual investor mix: Aware Super (Australia’s second-largest superannuation fund), Horizons Ventures (Li Ka-shing’s family office), and the National Reconstruction Fund Corporation (Australian federal government). The government co-investment adds a strategic-interest overlay to any acquisition process — expect consultation requirements if an overseas buyer approaches.

Revenue was approximately $4.75 million at time of raise — the valuation is entirely based on clinical validation and platform potential, not current revenue.

Strategic acquirer thesis: Philips, GE Healthcare, Canon Medical, or Fujifilm — all have PACS/RIS infrastructure in APAC hospitals and would acquire Harrison.ai to add AI diagnostic capability to their existing hardware relationships. The government investor creates a soft national-interest review even without formal FIRB requirements.

---

Mirxes

Mirxes built GastroClear — a blood-based AI diagnostic test for gastric cancer, the first blood test of its kind to receive regulatory approval globally. This is not a theoretical product: it is commercially approved and deployed. Gastric cancer is the fifth most common cancer globally but disproportionately prevalent in Asia (Japan, Korea, and China have the world’s highest incidence rates) — which means Mirxes is building products for the specific populations that need them most.

The A*STAR spin-off heritage gives Mirxes access to the Genome Institute of Singapore’s research infrastructure. The $40 million royalty-backed financing from CBC Group’s R-Bridge Fund in January 2025, followed by an SGX IPO in May 2025, provides commercial validation and capital for geographic expansion into Japan, Korea, and China.

A multi-cancer early detection (MCED) platform is in development — the next product that would expand the addressable market significantly.

Strategic acquirer thesis: A global IVD company (Roche Diagnostics, Illumina, Guardant Health) acquiring a proprietary miRNA platform with commercial revenue and APAC regulatory traction. The gastric cancer focus is a strategic moat that Western IVD companies lack — no equivalent product exists in Western markets.

---

Doctor Anywhere

Doctor Anywhere serves 2.5 million+ users across Singapore, Malaysia, Thailand (acquired Doctor Raksa — Thailand’s largest telemed platform), Philippines, Vietnam, Indonesia, and India. IHH Healthcare — Malaysia’s largest hospital group with 80+ hospitals — is an existing strategic investor, having led an earlier funding round.

The AI layer covers symptom checking, virtual consult triage, chronic disease management with AI-driven medication adherence, and remote monitoring. The recent acquisition of Asian Healthcare Specialists (a Singapore specialist clinic network) signals Doctor Anywhere is building a hybrid digital-physical network rather than remaining purely telehealth.

Total funding of $195 million includes a $38.8 million Series C1 led by Novo Holdings (Novo Nordisk’s investment arm) — healthcare-specialist capital with a long holding horizon.

Strategic acquirer thesis: IHH Healthcare is the obvious acquirer, given its existing equity stake and the strategic alignment of adding telehealth to its 80+ hospital network. Ping An Good Doctor (China), Fullerton Health, or a regional insurer (AIA, Prudential APAC) building embedded healthcare are alternatives.

---

Deepwise Healthcare

Deepwise has been a quiet achiever in China’s AI radiology market: first AI medical device approval in China for breast cancer screening, multiple NMPA Class III approvals (the highest regulatory classification), and deployment in 500+ Chinese hospitals across breast cancer, stroke, fracture, prostate cancer, and lung cancer detection.

The $68.6 million Series D in February 2025 (co-led by Legend Capital and Hangzhou Urban Industrial Fund) was the largest standalone funding round for a Chinese AI diagnostic company in that period. NMPA Class III approvals are genuinely hard to obtain — the Chinese regulatory process for AI medical devices is rigorous and the certification period is multi-year, creating a meaningful acquisition moat.

Strategic acquirer thesis: United Imaging Healthcare or Shanghai Mindray (Chinese medtech leaders) would acquire for AI diagnostic network coverage on top of existing hardware. Siemens Healthineers or Philips seeking a China market entry via an AI layer on top of existing hardware relationships in Chinese hospitals.

---

Nalagenetics

Nalagenetics is the smallest and earliest-stage company on this list, but it occupies a category that is structurally underserved: pharmacogenomics AI for Asian genetic variants. The company, spun out of A*STAR’s Genome Institute of Singapore, profiles patients’ genetic variations to predict medication response and adverse drug reactions.

The clinical significance in APAC is direct: Asian populations carry genetic variants at different frequencies than European populations, meaning that drug dosing guidance derived from Western clinical data is frequently suboptimal or potentially harmful. Nalagenetics’ models are specifically trained on Asian genomes.

Strategic acquirer thesis: A global diagnostics company (Roche, QIAGEN, Illumina) building APAC-specific clinical genomics capability, or a large hospital system (NUH, Tan Tock Seng, Tokyo Medical University Hospital network) wanting to embed pharmacogenomics AI into clinical workflows. At $13.6 million raised, this is an early-stage acqui-hire or seed-stage strategic investment target rather than a full M&A event.

---

Comparison Matrix

---

M&A Transaction Log

---

Valuation Benchmarks

---

APAC-Specific Dynamics

Japan as the highest-urgency market: Japan has the world’s most aged demographic and its own dedicated regulatory fast-track — the PMDA Sakigake designation for breakthrough AI medical devices. SoftBank’s formation of SB Tempus (a JV with Tempus AI) specifically for Japan precision oncology signals that Japan’s largest technology-telecom conglomerate views AI health as a structural priority. Fujifilm (imaging), Olympus (endoscopy), and Canon Medical (diagnostics) are Japanese medtech companies that have the revenue and the M&A appetite to acquire APAC AI health capability. No major AI health company has been bought outright by a Japanese acquirer yet — the NVIDIA-VinBrain template suggests the next APAC acquisition may come from a non-obvious buyer.

Korea’s dual role — acquirable and acquirer: Korea has approved 157 AI medical devices as of 2025 (up from 62 in 2023), the fastest regulatory throughput in APAC. Lunit and Seers are the two most prominent publicly-listed AI diagnostic companies in APAC. The Rznomics–Eli Lilly deal shows that Korean biotech platforms with novel science command global pharma pricing. Korean companies are simultaneously building products ready for export and building sufficient scale to attract premium acquisitions.

The biobucks pattern is APAC-specific: Unlike the US, where healthtech M&A is frequently outright acquisition, the APAC AI health deal structure strongly favors collaboration agreements with milestone payments. XtalPi ($6 billion structure, $51 million upfront), Rznomics ($1.3 billion structure, undisclosed upfront) — both are deals where the headline number is largely theoretical. M&A advisors working in APAC AI health must model milestone probability distributions and risk-adjust headline valuations, which requires both scientific judgment and financial modeling that few generalist advisors can provide.

Singapore as regulatory conduit: Singapore’s MAS/MOH regulatory framework, English-language legal system, and position as ASEAN’s corporate HQ make it the easiest APAC jurisdiction in which to close an AI health deal. Mirxes, Doctor Anywhere, and Nalagenetics are all Singapore-domiciled. Companies seeking to acquire APAC health AI capability frequently structure transactions through Singapore entities regardless of where the underlying technology is developed.

---

If your company operates in AI healthcare in APAC and you are considering a strategic process, contact us for a confidential conversation.