VCBeat, the Chinese healthcare-investment media outlet, published an on-record investor take on Chinese brain-computer interfaces on 2 August 2026. The piece is anchored by Jiang Donghui, partner at Delian Capital, a Chinese VC that has taken positions in Mindtrix (visual reconstruction) and co-led BCI-Sonics (transcranial ultrasound). The piece frames the sector as at risk of a bubble cycle before maturity, and Jiang offers investors a working matrix to sort tech-leading operators from concept plays before the compression corrects.
The financing number underneath the caution is the reason it matters. Per IT Juzi data cited in the piece, Chinese BCI companies completed 104 financing rounds totalling RMB 9.37 billion between 2013 and May 2026. First-half 2026 alone accounted for approximately 60 per cent of the thirteen-year total. Named single rounds include BrainCo’s early-2026 Series B+ at roughly RMB 2 billion (post-money $1.3 billion, the largest single round in Chinese BCI history), StairMed’s March 2026 strategic round of RMB 500 million led by Alibaba with Tencent participating (the first BCI investment by either), and Gestalt Technology’s cumulative RMB 570 million across an angel and Angel+ round in six months.
That is the compression Jiang is calling out.
Jiang’s four-quadrant matrix
Jiang gives investors a two-axis framework. Horizontal axis is demand space. Vertical axis is technological advancement. She discusses three of the four quadrants.
Upper right, technological leadership plus clear demand. Jiang treats this as the only ideal investment quadrant. Named examples: visual reconstruction (Delian’s Mindtrix position, 43 million blind and 295 million with moderate-to-severe vision impairment globally per WHO), and motor function reconstruction for spinal cord injury (Neuracle’s NEO, first commercial prescription issued in Shanghai on 13 July 2026 with the implantation performed at Huashan Hospital, which we covered separately).
Upper left, tech-leading but demand not yet materialised. Jiang’s example: ultrasound BCI, no Class III medical device approval anywhere globally to date. Jiang’s warning: “No matter how advanced the technology is, if it fails to translate into actual demand and lacks technological convergence, caution is warranted.”
Lower right, high demand but low technical barrier. Non-invasive neuromodulation for sleep disorders and depression, consumer EEG. Jiang: “This sector is somewhat mixed in quality. If you launch a mediocre product, consumers’ curiosity may drive them to give it a try; however, to truly address practical problems, you need to deliver a product that scores 80 to 90 out of 100.”
Three-path breakdown
The piece maps ten named Chinese BCI operators across the three signal-acquisition paths, plus Neuralink as international benchmark.
Semi-invasive: Neuracle (NEO, world’s first Class III approval for an invasive BCI from NMPA on 13 March 2026, epidural approach with 2mm skull groove, STAR Market IPO submitted June 2026, medical insurance code obtained within 48 hours of approval, 28 months from first clinical trial to approval). NeuCyber (Beijing Institute for Brain Disorders and Brain-like Intelligence spinout, Beinao No.1 system, 100+ channels, fully wireless, approximately 30 human implants by June 2026, 70,000+ hours safe operation, first patient 15+ months implanted). Neuracle is also positioned in the invasive category in a later paragraph because the piece treats its December 2025 fully-implantable-wireless-with-integrated-battery product separately, that product reportedly achieved 5.2 bits per second decoding at five days post-surgery.
Fully invasive: Neuralink (referenced as international benchmark, plans 3,000 electrodes in 2026 rising to 25,000+ by 2028), StairMed (256-channel wireless system launched December 2025, ultra-flexible electrodes at one-fifth the size of Neuralink’s, RMB 500 million Series B March 2026 Alibaba-led with Tencent), We-Linking (medical-grade fully-implantable systems, high-density flexible electrodes and specialised neuroelectrophysiology chips).
Non-invasive: BrainCo (smart bionic hands, bionic legs, attention training, sleep devices, revenue at scale in consumer market, RMB 2 billion Series B+ early 2026 at $1.3 billion post-money, IPO filed, competing with NeuroBooster to be “first listed BCI company”). BrainUp (sleep and mental health, series B in May 2022, cumulative hundreds of millions of yuan raised, investors including Chongqing Venture Capital, Ocean Capital and the Jinan Biomedical Medical Fund of Funds). ZhenTec (BCI plus exoskeleton for rehabilitation). HuiChuang Medical (fNIRS, first BCI assessment outpatient clinic in Central China with Tongji Hospital of Huazhong University of Science and Technology). Pengrui Neuroscience (portable EEG, Class II medical device certification). Gestalt Technology (ultrasound BCI without craniotomy or implants, RMB 420 million Angel+ round 3 July 2026 led by Meridian Capital with Sequoia China, Lens Technology and C Capital participating).
Per China Academy of Information and Communications Technology data cited in the piece, 88 per cent of domestic Chinese BCI companies have chosen the non-invasive path, citing safety, lower entry barriers and faster commercialisation. Jiang’s read: this path is “visibly crowded.”
Ultrasound as its own category is where Delian is placing capital. BCI-Sonics, founded July 2025, closed an angel round of hundreds of millions of RMB led by MPCi, followed by an angel+ round co-led by Delian and Tao Capital. BCI-Sonics claims a transcranial phase correction algorithm delivering 1.5mm transcranial focusing precision. Per Jiang’s investment commentary, capital is betting on full-chain capability from algorithms to transducers and from prototypes to regulatory registration, rather than on the fused “AI + Brain-Computer Interface” narrative.
US-China competitive framing
Jiang: “The gap between China and the United States is narrowing. While China may lag by three to five years in certain niche sectors, it has taken a leading or even forefront position in some technological pathways. However, claiming that China is comprehensively ahead would be an exaggeration. More accurately, China excels in rapid application translation and large-scale implementation, but there is still room for improvement in foundational, original technologies.”
The piece’s own summary line: “The world’s first Class III certification for an invasive brain-computer interface was not awarded in Silicon Valley, but in Shanghai.”
Cross-industry capital and exit pathways
Two to three years ago Chinese BCI capital came primarily from specialist healthcare funds. After the 2025 policy wave (Beijing and Shanghai city-level support in January, NHSA separate BCI project category in March, seven-ministry joint policy document in July, BCI written into the 15th Five-Year Plan by year-end), AI-focused and tech-focused funds surged in.
Jiang: “Rapid consensus among investors does not mean that everyone’s understanding is aligned; the fundamental logic for some capital has not been established.”
Jiang’s exit-path taxonomy: platform companies suit IPO (BrainCo in queue, PINS Medical preparing). Single-product or single-indication companies suit acquisition by majors (Medtronic and Boston Scientific are named as actively acquiring neurotechnology firms). Excessive-technology-cycle assets suit partial exit through S funds or secondary share transfers.
BCI-specific risk framing from Jiang: “The core risk of AI lies in its commercialization pathway, the core risk of innovative drugs lies in clinical data, and the core risk of brain-computer interfaces lies in the failure to translate technology into viable products.”
The “AI + Brain-Computer Interface” narrative pushback
Jiang holds a distinct counter-position on the “AI + Brain-Computer Interface” investment narrative that has gathered pace this year. Her line: “AI is now a foundational tool that every enterprise will utilize. But does the adoption of large language models by brain-computer interface companies necessarily enhance their algorithmic capabilities? Not necessarily. It is not a one-size-fits-all application.”
Her reason: BCI algorithms are highly customised. Motor decoding, language decoding and visual reconstruction each require distinct algorithmic logic. AI’s contribution sits at the foundational-tool level rather than at the commercialisation-accelerator level. The contrast Jiang draws: AI large-model companies pursue generality and scalability, but BCI algorithmic demands are distinctly non-generalisable, requiring customised development for each indication and each signal type.
What this piece is and is not
This is one investor’s on-record analytical framework, published in a Chinese healthcare-investment outlet on the day the Chinese BCI concept-stock rally week ended. It is one investor’s framework, not a market-wide consensus or a securities-firm downgrade.
What it adds to the July arc we covered separately (Neuracle first commercial prescription, StairMed return-to-work, NeuroXess xiangqi demonstration, Wang Ning hackathon, 28 July HK stocks rally) is a named Chinese investor voice explicitly warning that the current financing pace runs ahead of the sector’s translation-to-viable-product timeline, and offering a working matrix to sort operators before the compression corrects. The 60 per cent H1 2026 concentration figure is the strongest single quantification of the pace concern.
The near-term question is whether other Chinese analysts and securities firms echo the bubble-cycle language over the next two to four weeks. If they do, the July concept-stock rally is exposed to correction. If not, this stays a Delian-Capital-only view.