Where is there still room in neurotech?

Is neurotech actually crowded right now?

No. Neurotech looks crowded today because a few companies attract huge rounds and huge attention, while most of the market still has relatively few funded competitors and very little commercial penetration.

Neurofounders’ latest first-half 2026 review tracked 363 neurotech startups and initially found $2.64 billion across 81 funding events. After finding one additional convertible note, it updated the total to $2.66 billion across 82 events. Only about 22% of the companies had raised during the period, and nearly four in five had not.

More importantly, the money was concentrated at the top. Just 13 rounds of at least $50 million represented roughly three quarters of disclosed funding, while the median round was about $7.9 million. In BCI specifically, Merge Labs raised $252 million and Science Corporation raised $230 million. Together, those two rounds accounted for around 70% of the category’s $690 million total.

Neuromodulation looks healthier. Roughly $791 million went into the category, spread across companies working on brain stimulation, spinal stimulation, peripheral nerves and focused ultrasound. Diagnostics and assessment went the other way: only seven of 56 tracked companies raised in the first half, for about $48 million combined. Epitel then raised $26 million for remote EEG, and Hemispheric came out of stealth with a $52 million seed round, so activity has picked up lately without making the category remotely crowded.

Consumer neurotech provides another useful warning about headline numbers. Roughly $357 million was invested in the category during the first half, but BrainCo and Temple represented about 95% of it. Remove those two deals and the apparent funding wave gets small very quickly.

So we should be careful with the word “crowded.” There are several crowded financing races inside neurotech. There are far fewer mature product markets where startups are already fighting over the same paying customers.

Neurotech category	First-half funding	What the total hides	Our read
Neuromodulation	~$791M	Capital spread across many approaches	Broadest commercial opening
BCI	~$690M	~70% came from Merge Labs and Science	Crowded at the platform level
Consumer neurotech	~$357M	~95% came from BrainCo and Temple	Much thinner than it looks
Neuroimaging	~$125M	More evenly distributed	Selective room
Diagnostics & assessment	~$48M	Only 7 of 56 tracked companies raised	Still underbuilt

Has neurotech actually crossed from demos into real products?

Yes, parts of neurotech are becoming real medical markets now, although commercial maturity varies wildly depending on the technology.

The biggest change came from China. Neuracle Medical Technology’s NEO system received Chinese market approval for people with tetraplegia caused by cervical spinal-cord injury. The implant reads cortical activity and lets the patient control a pneumatic glove. Nature Biotechnology described the approval as the world’s first commercial authorization for an invasive BCI.

The milestone has already moved beyond paperwork. A patient subsequently received the first commercially prescribed NEO implant at Huashan Hospital in Shanghai. Until then, essentially every next-generation movement and communication BCI we discuss in the West was still being used inside a clinical study.

We can see the same transition further along the commercialization curve with NeuroPace. Its RNS epilepsy system generated $22.5 million of revenue in the second quarter of 2026, up 21.3% from the previous year, and the company raised its full-year revenue guidance to $99.5 million to $101.5 million. NeuroPace is still loss-making, but this is already a real device business growing at more than 20% in its core indication.

Precision Neuroscience sits somewhere between those examples. Its Layer 7 cortical interface has FDA clearance for recording, monitoring and stimulation for up to 30 days, has been used in more than 95 patients and is now deployed across more than 15 hospital partners. Its fully implantable BCI remains in development.

Those three cases show how uneven neurotech currently is. Commercial brain interfaces exist. Chronic communication BCIs are still experimental in most countries. Established therapeutic implants already generate meaningful revenue. A startup founder therefore needs to ask which part of the clinical stack has actually matured rather than treating “neurotech” as one market.

Is there still room for another Neuralink-style brain implant?

Very little room remains for a generic Neuralink-style BCI whose main pitch is simply more electrodes, more bandwidth and thought-controlled computers.

Neuralink had publicly announced 21 implanted participants by early 2026 after raising $650 million in its Series E. Synchron is currently running clinical studies around an entirely different implantation route, placing the Stentrode through the jugular vein rather than opening the skull. Paradromics recently implanted the first participant in its long-term Connect-One study, which is testing speech restoration and computer control. Precision already has more than 95 patients of experience with its cortical surface array.

And the race is no longer confined to U.S. clinical trials. Neuracle has become the first company to move an invasive BCI into commercial clinical use in China.

These companies cover most of the obvious architectural positions. Neuralink and Paradromics pursue high-bandwidth intracortical recording. Precision uses a high-density surface array. Synchron comes through a blood vessel. Neuracle uses an epidural approach and already has a narrowly defined commercial indication.

A startup can still win in implanted BCIs, but “we built a better brain chip” is becoming a weak reason to start one. We would want a much sharper wedge: a procedure that is dramatically easier, an indication ignored by the existing companies, a sensory interface, a therapeutic application or an interface that fits into a clinical procedure physicians already perform.

The interesting whitespace has shifted toward deciding exactly what the interface should do for a patient.

If you want more recent data on this point, please see our latest neurotechnology market report.

Is making brain implants easier to implant more valuable than adding more electrodes?

For most new BCI companies today, reducing surgical burden looks more valuable than simply pushing electrode count higher.

Synchron makes the trade-off easy to see. Its Stentrode does not sit directly inside cortical tissue, so it gives up some of the signal resolution available to penetrating arrays. In return, physicians deliver the device through the jugular vein in a procedure Synchron says takes around two hours, with most trial participants going home the next day.

Precision is taking another route. Its Layer 7 array sits on the cortical surface, and the company has deliberately built its clinical program around procedures neurosurgeons already perform. Its partnership with Medtronic goes further by integrating Layer 7 with Medtronic’s StealthStation surgical-navigation platform. That is a much more commercially useful milestone than another laboratory record for channel count because it deals directly with how the technology reaches operating rooms.

Neuracle provides the strongest real-world validation of this direction so far. Its commercially approved NEO device sits outside the dura rather than penetrating the cortex. China has therefore reached commercialization first with a relatively low-channel, narrowly targeted interface rather than the highest-bandwidth implant.

There will still be applications where enormous neural bandwidth matters. Rich speech decoding, dexterous motor control and future sensory interfaces may need it. But once a device has enough information to perform its job, lowering operating time, tissue damage, infection risk, recovery time and specialist burden starts to matter a lot more.

That creates room for startups working on delivery systems, flexible interfaces, wireless power, packaging, minimally invasive implantation and procedure-specific neural devices. Neurotech companies sometimes talk as though the whole race happens at the electrode. Increasingly, the operating room is deciding who can scale.

Is speech restoration the best open market in brain-computer interfaces?

Yes. Speech restoration currently has the clearest combination of extraordinary technical progress, obvious patient value and no established commercial winner.

A recent Nature Medicine study is the strongest evidence we have seen that speech BCIs can work outside a laboratory. A man with ALS used an intracortical BCI at home for more than 3,800 hours over 19 months. He produced 183,060 sentences, nearly 2 million words, at an average of 56 words per minute. He rated 92% of sentences as at least mostly correct, while controlled testing exceeded 99% word accuracy using a 125,000-word vocabulary.

The system was useful enough for the participant to communicate with family and colleagues, browse the internet and continue working full-time. Earlier research from the same broader group had already shown instantaneous synthesized speech with changes in intonation. Another Berkeley-UCSF team has demonstrated streaming brain-to-voice output with very low latency.

Commercial companies are now following the academic results. Paradromics’ first long-term Connexus implant is being studied specifically for speech restoration and computer control. Neuralink has also received FDA Breakthrough Device designation for a future speech-restoration program.

The remaining gap is easy to understand. The 3,800-hour system still involved implanted arrays connected to external research equipment, and care partners helped connect the hardware and start the system. Researchers have shown that the decoding can work for everyday communication. Nobody has yet packaged that performance into a wireless medical product that thousands of people can receive, maintain and pay for.

That makes speech BCI one of the rare neurotech markets where the science has moved surprisingly far ahead of the product. There is real room here, although the capital, clinical evidence and hardware execution required will eliminate most startups before they begin.

If you want more recent data on this point, please see our latest neurotechnology market report.

Are therapeutic neurodevices a bigger opportunity than communication BCIs?

Right now, therapeutic neurodevices offer a larger and more proven business opportunity than communication BCIs.

Medtronic’s adaptive deep-brain stimulation system shows how far therapeutic interfaces have already gone. BrainSense Adaptive DBS listens to brain activity and changes stimulation automatically in people with Parkinson’s disease. Medtronic had already implanted more than 40,000 patients with Percept devices when the adaptive feature received U.S. approval, and the company says it has treated more than 185,000 people with DBS since 1987.

NeuroPace gives us an even cleaner startup comparison. More than 8,000 patients have received its responsive neurostimulation system for epilepsy. The company now has over 35,000 patient-years of clinical experience and more than 26 million intracranial EEG recordings. Its latest quarterly RNS revenue grew 21.3% year over year, and gross margin remained above 80%.

Compare that with high-bandwidth communication BCIs, where even the best-known commercial programs are still measured in dozens of implanted trial participants. The technical ambition may be higher, but the business is much earlier.

The next opportunity is making stimulation more adaptive. Instead of programming a device during occasional clinic visits and hoping that one setting works throughout the day, a closed-loop system can measure the nervous system, detect a useful state and change therapy automatically.

Parkinson’s and epilepsy have already proved that loop can work. Depression, chronic pain, stroke rehabilitation, spinal-cord injury and other neurological conditions give startups much more territory to explore. If we were choosing between another general-purpose computer-control BCI and a closed-loop device aimed at one serious neurological disease, we would currently take the therapeutic device.

Which neurological diseases still have the clearest neurotech openings?

Epilepsy, depression, stroke and spinal-cord injury currently offer some of the clearest openings because each combines a large treatment gap with measurable neurological signals or functions that a device can influence.

Epilepsy is already commercially validated. NeuroPace’s core RNS business is growing above 20%, yet the company still describes hundreds of thousands of Americans as potential candidates within drug-resistant focal epilepsy. Expansion is not automatic: the FDA recently asked NeuroPace for more clinical information before approving an extension into idiopathic generalized epilepsy. Good neuroscience does not erase the regulatory burden.

Depression is developing along several competing technical paths at once. Prescription home stimulation has reached FDA approval, TMS is already established, and companies are testing more targeted implanted systems. There is still no dominant architecture for delivering personalized, convenient neuromodulation to the enormous population that does not respond adequately to medication.

Stroke rehabilitation has become much more interesting lately. A Nature Medicine feasibility study implanted cervical epidural spinal-cord stimulators in seven people with chronic post-stroke arm weakness. With stimulation on, average strength improved by 32% and the Fugl-Meyer motor score improved by 5.6 points. Three of seven participants recovered additional hand or finger movement. Seven people is far too small to settle efficacy, but the size of the functional change makes the approach worth following.

Spinal-cord injury is further ahead commercially because non-invasive stimulation has already reached clinics and homes. Parkinson’s, by comparison, is a much more mature implant market with Medtronic and other large device companies already deeply entrenched.

Condition	What neurotech has already proved	How much room remains	Our view
Epilepsy	Closed-loop implants can become a growing commercial business	High inside underpenetrated populations and new indications	Strong
Depression	Home stimulation can clear FDA review	Very high because no approach dominates	Strong
Stroke	Early spinal stimulation can improve chronic arm function	Very high, but evidence remains early	High upside, high risk
Spinal-cord injury	External stimulation is already being commercialized	High across home use, implants and rehabilitation	Strong
Parkinson’s disease	Adaptive DBS is already commercial at scale	More room around software and targeting than basic DBS hardware	Selective

Is spinal-cord stimulation the most overlooked opportunity in neurotech?

Spinal-cord stimulation is one of the strongest underbuilt neurotech markets today because it can restore useful function without requiring us to decode every intended movement directly from the brain.

ONWARD Medical provides the clearest commercial example. Its non-invasive ARC-EX system received U.S. clearance for improving hand strength and sensation after chronic cervical spinal-cord injury and later received clearance for home use. ONWARD reported 70 ARC-EX units supplied in the first quarter of 2026, with penetration of more than 80 U.S. clinics and more than 100 clinics globally.

Those numbers are still tiny compared with mainstream medical devices, which is exactly what makes the category interesting. A clinically validated product has reached the market before the market itself has become large.

The opportunity is also expanding beyond spinal-cord injury. The seven-person stroke study discussed above found immediate gains in strength and motor function from cervical epidural stimulation after years of chronic impairment. Participants received fewer than nine hours of movement-based training over four weeks, yet average motor scores remained improved at the end of the study.

The underlying idea is powerful. Many patients have damaged neural pathways rather than completely absent ones. Stimulation can make the surviving circuits easier to recruit, allowing the patient’s own nervous system to generate movements that were previously too weak or poorly coordinated.

That opens several startup layers: stimulation hardware, implantation, patient-specific programming, rehabilitation software, sensors that measure recovery, home treatment and systems that combine cortical intent with spinal stimulation. Neurorehabilitation could ultimately produce a larger patient market than direct computer-control BCIs because the goal is to improve the body itself rather than route around it.

If you want more recent data on this point, please see our latest neurotechnology market report.

Is vision restoration the most underrated neuroprosthetics market?

Vision restoration is currently one of the most underrated neurotech opportunities because functional restoration has already been shown in a reasonably sized multicenter study.

Science Corporation’s PRIMA retinal implant was tested in 38 people with geographic atrophy caused by advanced dry age-related macular degeneration across 17 sites in five countries. Among the patients evaluable after 12 months, about 80% achieved a clinically meaningful improvement in visual acuity. The average improvement reported in the New England Journal of Medicine study was 25.5 letters, or more than five lines on a standard eye chart.

More importantly, patients were able to read letters, numbers and words using prosthetic central vision. That gives PRIMA an unusually tangible endpoint for neurotech. We do not need to argue about whether a change in a neural biomarker is meaningful when a previously functionally blind person can read again.

Science has since raised $230 million to push PRIMA toward commercialization and expand clinical work into conditions including Stargardt disease and retinitis pigmentosa. The company has submitted applications in both Europe and the United States, although PRIMA remains investigational and has not yet completed the regulatory path to broad commercial availability.

Vision also has a structural advantage over a general-purpose BCI. The patient population is identifiable, ophthalmologists already perform retinal surgery, visual acuity provides a familiar regulatory endpoint and the benefit is immediately understandable.

That combination is rare in neurotech. We would pay much more attention to disease-specific sensory prostheses than the volume of press coverage currently suggests.

Can at-home neurostimulation become a real depression market?

Yes. At-home neurostimulation for depression has moved into real medicine, and the market is currently open enough for several approaches to compete.

Flow Neuroscience’s FL-100 received FDA premarket approval for adults with moderate-to-severe major depressive disorder who are not considered treatment-refractory to medication. The pivotal study included 174 participants and was run remotely. At ten weeks, the active-treatment group improved by 11.3 points on the MADRS depression scale versus 7.7 points for sham treatment. Clinical response reached 63% versus 31.6% for sham, while remission was 57.5% versus 29.4%.

Neurolief has taken a different route with Proliv Rx. The FDA approved the device as an adjunctive treatment for adults with major depression who had not improved adequately after at least one antidepressant. Proliv can also be used at home.

That gives us two prescription home neuromodulation systems reaching the U.S. market through different stimulation approaches. At the more invasive end, Motif Neurotech has received permission to start human testing of a small depression implant designed to sit in the skull above the dura and deliver targeted electrical stimulation without placing electrodes directly into brain tissue.

The commercial opening comes from the gap between antidepressant pills and clinic-heavy procedures such as TMS or implanted DBS. A device that patients can use repeatedly at home changes the economics and convenience of treatment.

We still need much more evidence on long-term adherence, durability of response, reimbursement and which patients respond to which stimulation pattern. Those unknowns make the market risky, but they also explain why a winner has not already emerged.

Is focused ultrasound still too early for a neurotech startup?

Focused ultrasound is early, but the technology has become important enough that we would no longer dismiss it as speculative neurotech.

The attraction is straightforward: ultrasound can reach deeper brain structures without putting an electrode into them. Electrical stimulation delivered through the scalp has limited spatial precision at depth, while conventional DBS requires surgery. Focused ultrasound could eventually occupy a valuable middle ground.

The field has started attracting serious company formation and capital. Merge Labs raised $252 million to pursue a neural-interface approach involving ultrasound and biological targeting. Gestala is also developing ultrasound-based systems, initially around applications such as chronic pain. Neurofounders’ funding data shows neuromodulation attracting more capital than any other major neurotech category in the first half of 2026, with focused-ultrasound companies included in that flow.

The clinical evidence for low-intensity neuromodulation remains much weaker than the funding. Researchers are studying targets related to depression, addiction and other psychiatric conditions, but sample sizes are generally small and stimulation parameters are far from standardized. The skull also distorts acoustic energy, making individualized targeting and dose control difficult.

We would separate focused-ultrasound therapy from the much more speculative idea of using ultrasound as a rich non-invasive BCI for reading complex thoughts. The therapeutic version has a plausible clinical path today. High-bandwidth non-invasive neural decoding is much further away.

For founders, focused ultrasound offers genuine technical whitespace around transducers, targeting, imaging, treatment planning, acoustic modeling and disease-specific protocols. It is one of neurotech’s highest-upside areas, but certainly not one of its safest.

If you want more recent data on this point, please see our latest neurotechnology market report.

Is neurodiagnostics the easier way to build a neurotech company?

Yes. Neurodiagnostics currently gives startups a cleaner route into the market than implants because measuring neurological activity usually carries less clinical risk while still solving expensive problems.

Beacon Biosignals is a useful example. The company has now raised more than $132 million in total, including a Series B expanded to $97 million, around a platform for at-home EEG and quantitative neurophysiology. Beacon says it has analyzed more than 2 million hours of EEG and supported more than 40 clinical trials. Harmony Biosciences is using Beacon’s at-home EEG technology in Phase 3 studies for narcolepsy and idiopathic hypersomnia.

Epitel attacks the same problem from another angle. The company recently raised $26 million to expand REMI, its FDA-cleared wireless EEG system for extended monitoring. The round is going toward commercial expansion rather than proving that EEG can work. Epitel’s core pitch is practical: neurological events such as seizures can be intermittent, while conventional EEG access is limited and short hospital recordings often miss them.

The more interesting software opportunity appears once a company owns longitudinal neurological data. NeuroPace has accumulated more than 26 million intracranial EEG recordings across over 35,000 patient-years. The FDA recently approved its ECoG Assistant, which uses that dataset to highlight intracranial recordings that may contain clinically important events. NeuroPace says the system was developed from more than 124,000 expert-labeled records and identifies ECoGs of interest with 93% sensitivity.

That is a much stronger data moat than a generic “AI for brain signals” pitch. The software is attached to a differentiated dataset, a regulated medical product and a decision a clinician already needs to make.

This is where we see the real room in neurodiagnostics: seizure detection, neurological trial endpoints, treatment-response monitoring, patient selection and continuous measurement between visits. Another dashboard showing a patient an abstract brain score would be far less interesting.

So where is there still room in neurotech?

There is plenty of room in neurotech today, but we would build around a specific neurological problem rather than another general-purpose brain interface.

The latest evidence makes the hierarchy clearer than it was even recently. Neuromodulation attracted roughly $791 million in first-half financing, more than any other core neurotech category, and the money was distributed across many companies rather than one or two mega-rounds. NeuroPace’s RNS business is growing above 20%. ONWARD is moving spinal stimulation through clinics and into homes. Flow and Neurolief have put prescription depression stimulation on a home-use regulatory path. Remote EEG companies are raising money to expand commercially. Speech BCI research has reached thousands of hours of independent home use. China has already moved an invasive motor BCI into commercial clinical practice.

We would put closed-loop therapeutic neuromodulation at the top of the list because the commercial model is already proven while many diseases remain poorly served. Spinal stimulation and neurorehabilitation come close behind because the clinical evidence is improving quickly and competition remains relatively thin. Neurodiagnostics offers a simpler startup wedge, particularly when neurological measurements change drug-development or treatment decisions.

Speech restoration still has exceptional upside, but the capital and clinical requirements make it a harder company to build. Vision and sensory prostheses look underappreciated because the clinical endpoints are concrete and recent results are unusually strong. Focused ultrasound deserves serious attention, although we would expect much more technical and clinical failure before the category settles.

We would be much less excited about another generic Neuralink competitor or another consumer headset promising better focus. Both markets already attract disproportionate attention relative to the evidence that a new entrant has a defensible place.

The broad opportunity in neurotech has become more practical. We already know that nervous-system signals can be recorded, decoded and changed. The companies worth building now are the ones that turn those capabilities into treatments or functions that patients will use repeatedly, physicians will prescribe and healthcare systems can actually deliver.

Open neurotech market	Why there is still room	Difficulty	Our judgment
Closed-loop neuromodulation	Commercial proof exists, but most diseases remain underpenetrated	High	Best overall opportunity
Spinal stimulation & neurorehabilitation	Strong new functional results with few scaled competitors	High	Most underbuilt
Neurodiagnostics & remote monitoring	Clear clinical bottlenecks and easier regulatory path	Medium	Best startup wedge
Speech restoration BCI	Performance is now remarkable but products remain experimental	Very high	Exceptional upside
At-home psychiatric neurostimulation	Regulatory validation has arrived before market consolidation	Medium-high	Very attractive
Vision & sensory prostheses	Functional restoration is becoming measurable	Very high	Underappreciated
Minimally invasive neural interfaces	Surgical burden still limits adoption	Very high	Important platform opportunity
Focused-ultrasound neuromodulation	Deep targeting without implants could unlock new treatments	Very high	High upside, still early
Generic high-bandwidth BCI	Several heavily financed architectures already compete	Extreme	Poor entry point
Generic consumer brain wearables	Weak outcomes, low switching costs and thin funding outside a few companies	Medium	Generally avoid

If you want more recent data on this point, please see our latest neurotechnology market report.

Who is the author of this content?

NEW MARKET PITCH TEAM

We track new markets so founders and investors can move faster

We build living "market pitch" documents for emerging markets: AI, synthetic biology, new proteins, and more. Instead of outdated PDFs or hallucinated LLM answers, our clients get a clean, visual, always-updated view of what's really happening: key players, deals, regulations, and signals that matter. Learn more about us.