China's Brain Chip Market Outruns Its Patients
The Boring Implant
Huashan Hospital in Shanghai implanted a Neuracle NEO device in a patient with decade-old spinal cord injury on July 14, the first commercial, not experimental, use of an invasive brain-computer interface in China. The device itself is the less interesting half of the story, and it is worth being honest about why. NEO carries eight electrodes on a subdural array, meaning it sits on the surface of the brain, tucked under the dura membrane, rather than piercing the cortex the way Neuralink's threads do. Fewer electrodes and no penetration means a coarser signal: you can read broad intent, open hand, close hand, more reliably than fine finger-by-finger control. What you get back is a device that heals like a drain placed under the skull rather than a craniotomy with wires threaded into grey matter, which is a real part of why regulators can wave it through faster and why surgeons can now offer it outside a trial. Neuracle has done this 36 times total, 32 of them in 2025 alone, per MIT Technology Review's reporting. That is a small, young case series for something now sold as a marketed product. The next part is where marketed gets ahead of proven.
Money Before Medicine
China's National Healthcare Security Administration assigned NEO a reimbursement code within 48 hours of the National Medical Products Administration's March 13 approval, according to Bird & Bird's regulatory analysis. Two days, for a code that in most systems takes months of price and outcomes review to attach to a device. Neuracle then filed for a Shanghai STAR Market IPO on June 11, seeking 2.5 billion RMB (about $345m), even though the implant itself produced zero revenue in 2025. Its actual 108 million RMB in sales, up 64 percent year on year, came from non-invasive EEG headsets, not the brain surgery product on the prospectus cover. Rivals read the same signal: StairMed in Shanghai raised 500 million RMB led by Alibaba, and Gestala in Chengdu closed a 150-million-RMB angel round for non-invasive BCI work. None of this is fraud. It is capital following an approval pathway and a reimbursement code, ahead of the outcomes data that normally justifies both. My read is that this inverts how a medical device market is supposed to form: proof, then pricing, then money. Here the pricing and the money arrived first, and the proof is still a 36-patient case series.
The same week, Skyroot's Vikram-1 reached 450km orbit from Sriharikota, a gate that either happens or does not: the rocket is up or it is in the sea. Neuracle's gates are softer, a code assigned, a prospectus filed, a round closed, none of which require the implant to work better than it did on patient 32. The real test arrives when provincial reimbursement talks start next year. Does the 2.5 billion RMB valuation survive contact with an actual price negotiation?
Sources
- MIT Technology Review -- NMPA's March 2026 commercial approval and device specs (technologyreview.com)
- South China Morning Post -- Huashan Hospital implant procedure (scmp.com)
- Bird & Bird -- 48-hour reimbursement code and regulatory pathway (twobirds.com)
- hellochinatech.com -- Neuracle STAR Market prospectus, zero implant revenue (hellochinatech.com)
- WION -- rival BCI funding and China-Neuralink comparison (wionews.com)
- Al Jazeera -- Skyroot's Vikram-1 orbital launch details (aljazeera.com)