A BGI Genomics visit is the closest thing Shenzhen has to a working museum of the life-sciences industry: a building shaped around a double helix on the coast at Dameisha, public galleries that walk through 3.8 billion years of evolution, and behind glass on the fifth floor the sequencing instruments that the group builds for itself rather than buys. BGI Genomics (300676.SZ) reported revenue of 3.71 billion yuan for 2025, down 4.18% year on year, while narrowing its net loss attributable to shareholders by 31.69%; the first quarter of 2026 turned revenue back to growth at 809 million yuan, up 20.35%. What a delegation can actually see is the part that matters — the exhibition floors, the viewing corridor above the sequencing hall, the multi-omics and AI platforms, and a meeting room where the company's own clinical and commercial teams take questions. That combination, a public science centre and a listed diagnostics business under one roof, is what makes the visit worth a morning.

Key takeaways

The programme runs at the BGI Time-Space Centre in Dameisha, Yantian District, Shenzhen. The standard format is half a day, either 09:00 to 12:00 or 14:00 to 17:00, and combines the exhibition floors, the viewing corridor above the sequencing hall, a themed session with the company's own teams and a closing discussion. The site opens 09:00 to 17:00 through the year, by appointment, and takes organised groups rather than individual walk-ins.

BGI Genomics is a diagnostics and precision-medicine business rather than a pure sequencing service shop. The listed company was established in 2010 and listed on the Shenzhen Stock Exchange's ChiNext board in 2017 under ticker 300676, while the wider organisation dates its own story to 1999 and China's one-per-cent contribution to the Human Genome Project, which is where the exhibition begins. By the end of 2025 the company was working with more than 1,000 medical institutions, disease-control centres and research institutes, and its infection assays had been used on more than 750,000 clinical patients.

The business is mid-transition, and that is the interesting part for a visitor. Revenue fell 4.18% in 2025 while the loss attributable to shareholders narrowed by 31.69%, and the first quarter of 2026 returned revenue to growth at 809 million yuan, up 20.35%, with the quarterly loss narrowed by 81.93%. Underneath those totals the model is moving from send-out testing, where samples leave the hospital, to in-hospital capability, where the laboratory, the instruments, the analysis software and the reporting are installed inside the hospital and the revenue recurs.

The international footprint is what overseas delegations come to test. By the end of 2025 the company had completed more than 154 technology-transfer projects across 51 countries, 38 of them in 18 countries during 2025 alone, covering NIFTY non-invasive prenatal testing, clinical exome sequencing, carrier screening, the Colotect colorectal cancer screening assay, hereditary tumour panels, HPV typing and pan-cancer assays.

The session is where the visit earns its time. Standard themes are the localised precision-medicine model — how a hospital stands up its own genomics laboratory with the SIRO solution — or the AI reporting stack, iGeneT Pro and ChatGeneT, that sits behind it. We arrange the programme end to end, including scheduling on the Chinese side and English interpretation on site.

Key visit details

Where the visit is held

At the BGI Time-Space Centre on the coast at Dameisha, in Yantian District on the eastern side of Shenzhen. The nearest metro is Dameisha station on line 8, roughly five minutes by car or twenty minutes on foot from the gate, and confirmed groups can park on site. The building is roughly an hour from Futian or Nanshan depending on traffic, which is the single most important planning constraint on the day.

Format and duration

Half a day, in the sequence the site works to: the exhibition floors, the fifth-floor viewing corridor above the sequencing hall, a themed session, then open questions. The outdoor field stations and the roof garden add about thirty minutes where the schedule allows. A full-day programme that adds a second company is possible, but the transfer out to Dameisha comes out of the programme rather than out of the day.

Who the programme suits

Five groups get proportionally more from it. Hospital groups and laboratory networks, because the SIRO model is a procurement decision they can evaluate in the room. Diagnostics distributors, because the transfer list tells them which assays can be registered and sold locally. Public-health and ministry delegations, because population-scale screening economics are the transferable lesson. Investors, because the segment mix explains the margin structure better than any slide. And pharmaceutical or biotech research teams, because the multi-omics services and the variant knowledge bases are the parts they would actually buy.

Booking, group size and site rules

Around 30 days is the practical lead time. The programme starts at 10 participants. Foreign visitors need their passport details registered with the site in advance, which is a longer process than for a domestic group, and photography is restricted in the laboratory areas. The sequencing hall is observed from the corridor; entry into the laboratory itself is at the host's discretion on the day and should not be assumed when planning the agenda.

Why a BGI Genomics visit is worth a delegation's day

Three questions bring most delegations here, and the site is unusually good at answering all three honestly. What has China actually industrialised in genomics, rather than demonstrated in a paper? What can be transferred to another country, and what stays domestic? And what does screening at population scale really cost to run?

The first is answered on the ground floor, in a single chart. One line falls — the cost of sequencing a genome. The other rises — the number of people who can be reached with it. The company calls it the scissors gap, and the whole visit is an argument about how fast the blades have closed. Non-invasive prenatal testing is the proof case: a test that existed as a laboratory result in one country became a population screening programme in several, and BGI was one of the companies that industrialised it and then carried it abroad.

The second is answered by the transfer list, which is more specific than most visitors expect. Eighteen countries took 38 transfer projects in 2025 alone. What moves is not a laboratory in a box but a defined package: the assay, the instrument, the analysis software, the training and the database, handed to a local partner who then runs it. What does not move is equally clear once you see it — the data infrastructure and the interpretation layer stay with the company.

The third is answered in the session rather than on the tour. Population screening is not expensive because sequencing is expensive; it is expensive because interpretation, counselling and follow-up are. That is precisely why the company has been building AI into the back end of its reports, and why the economics question is better put to the clinical team in the room than to the exhibition.

Inside the BGI Time-Space Centre

The exhibition floors: a cost curve and a one-per-cent story

The route opens with a wall that runs through 3.8 billion years of life, dinosaur reconstructions in the planting outside, and then the company's own history: bronze bricks replicated from the Millennium Monument in Beijing, the pipette tips used in 1999 when BGI researchers joined the Human Genome Project's one-per-cent contribution, and the scissors chart described above. Two details are worth slowing down for. The BGI tree carries project plaques from the company's work abroad, including the Fire Eye laboratories built during the pandemic and desert-greening projects. The BGI smile is a double helix built from twenty-three books, one for each chromosome pair, with four spine colours for the four bases.

The sequencing hall viewing corridor

On the fifth floor a ring corridor looks down into the sequencing laboratory through glass. This is the segment executives remember, because it is the only place on the route where the scale is legible: rows of instruments running continuously, and a guide explaining that the great majority of them are BGI-group machines rather than imported ones. MGI Tech, which manufactures the instruments, is a sister company under the same group rather than part of the listed entity — a distinction worth stating out loud during a visit, because it is the first thing an informed delegate will ask about.

The instrument range is the other thing to read properly. The E25 is a desktop sequencer built for a hospital laboratory that has no genomics department. The T7 is the high-throughput workhorse that carried pandemic-era volumes. The T20 is the top of the range, sized for around 50,000 whole genomes a year. What a delegation should look for is not the hardware but the flow: sample in at one end, interpreted report out at the other, with as few human hands in between as the laboratory can arrange.

The multi-omics and AI platforms

The platforms are shown in the themed session rather than on the route, and they are the most commercially loaded part of the visit. iGeneT Pro handles the analysis, variant filtering, classification and report drafting that used to sit with a small number of senior interpreters. ChatGeneT is the counselling-side tool that helps a clinician explain a result to a patient, with an education version used to train genetic counsellors. Both sit inside the BGI Zhihui medical system that the company sells to hospitals, and i99 extends the same logic into ongoing health management.

Behind them sits the data. The Wanxiang database organises what the company has accumulated across reproductive health, oncology, infection and chronic disease, and two variant knowledge bases — one for hereditary cancer, one for somatic variants and drug response — cover more than 200,000 variants across solid tumours and have been listed on the Shenzhen Data Exchange. For a delegation thinking about data as an asset rather than a by-product, this is the part to press on, and it is where the company is least like a conventional diagnostics business.

The outdoor field stations and the roof garden

Outside the main building the company keeps a field station with its multi-graft tomato tree, which carries several fruit species on one rootstock, and perennial rice that is planted once and harvested for years, plus a small animal enclosure and a roof garden of protected species. This is the segment the site builds for school groups. Executive delegations take it when the agenda touches agriculture or food, or when they want a lighter ending to a dense morning; it is easy to drop when they do not.

What the 2025 results actually show

The headline numbers are the ones a delegation should have before arriving, because they shape every question in the room. Revenue of 3.71 billion yuan, down 4.18% year on year. Net loss attributable to shareholders of 617 million yuan, narrowed by 31.69%; on an adjusted basis, a loss of 636 million yuan, narrowed by 30.82%. Gross margin of 37.59%, down 3.88 percentage points. Operating cash flow swung to an outflow of 63 million yuan. Headcount of 3,580, down 3.3%, with revenue per employee of about 1.035 million yuan.

The segment detail is where the story is. Precision medicine integrated solutions — the in-hospital model — brought in 1.59 billion yuan, up 19.45%, and now accounts for 42.84% of revenue. Reproductive health research and clinical services, the traditional send-out testing base, brought in 826 million yuan, down 28.06%. Multi-omics and synthesis services were flat at 668 million yuan, though single-cell sequencing within it grew by about 87%. Oncology and chronic disease services brought in 511 million yuan, down 18.28%. Infection prevention and control was the fastest-growing line at 95 million yuan, up 30.29%, with the PTseq product family roughly doubling.

Read together, the direction is unambiguous. The send-out testing business, exposed to procurement price pressure and to demographic headwinds in reproductive health, is shrinking. The in-hospital business, where the company installs capability and sells into it repeatedly, is growing. The company describes this as a shift from an independent clinical laboratory model to an in-vitro diagnostics model, and the SIRO solution is the vehicle: laboratory design, instruments and reagents, analysis software, technology transfer, training, database build and report interpretation, sold as one package. SIRO-48, released in the second half of 2025, extends it down to county hospitals and prefecture-level maternal and child health hospitals at mid-throughput.

The export model: technology transfer, not sample shipping

Most overseas delegations arrive assuming that a Chinese diagnostics company exports by taking samples from abroad and sequencing them at home. That is not the model, and seeing why is one of the more useful outcomes of the visit. The export unit is a transfer project: a partner in another country receives the assay, the instrument, the software, the training and the database, and runs the service locally. More than 154 such projects had been completed across 51 countries by the end of 2025, with 38 completed in 18 countries during the year itself.

The transferred product list is longer than most visitors expect: NIFTY for non-invasive prenatal testing, clinical exome sequencing, carrier screening, the Colotect colorectal cancer screening assay, hereditary tumour panels, HPV typing, and pan-cancer, lung, gastric and colorectal assays. The company has converged on a two-engine structure abroad — reproductive health plus infection control — because those are the two areas where a local health system will fund screening at scale.

Regionally, the company reported Asia and Oceania revenue up 16.46% in 2025, steady growth in South East Asia, and a doubling of revenue in Latin America and South Asia from a small base. On the domestic side it counts more than 1,400 joint laboratories with Chinese medical institutions and cooperation with more than 1,000 medical institutions, disease-control centres and research institutes worldwide, with its PMseq and PTseq infection assays used on more than 750,000 patients. For a delegation from a middle-income health system, those two numbers — 1,400 joint labs and 51 countries — are the ones that translate directly into a plan.

What different teams take away

For hospital groups and laboratory networks — whether a genomics capability can be stood up inside an existing hospital, what SIRO actually includes, how long the technology transfer takes, and what the reporting workflow looks like once the instruments are installed.

For diagnostics distributors — which assays on the list can be registered in their market, what the instrument footprint is at each throughput tier, and where the company will and will not go.

For public-health and ministry delegations — the economics of screening an entire birth cohort or an entire age band, how the cost curve has moved, and what the counselling burden looks like at the end of the pipeline.

For investors — why a falling revenue line and a narrowing loss are compatible, which segment carries the margin, and whether the data assets are a business or a story.

For pharmaceutical and biotech research teams — the multi-omics service range, the single-cell capability, and the variant knowledge bases as a research input rather than a clinical product. The same half-day should produce a different debrief for each of those five groups.

Sample visit programme

This is the shape of a standard morning, and it is the order the site works to. The afternoon window follows the same sequence between 14:00 and 17:00.

Segment Time What to expect
Arrival, registration and the evolution corridor 09:00–09:30 Sign-in with the visitor list submitted in advance, then the first galleries: the 3.8-billion-year wall and the company history that frames everything after it.
Exhibition floors 09:30–10:15 The scissors chart, the Human Genome Project material, the BGI tree and the BGI smile, with the guides translating the technical story into a cost and scale story.
Sequencing hall viewing corridor 10:15–11:00 Fifth-floor ring corridor above the sequencing laboratory: the instrument range from the desktop E25 to the T20, and what sample-in, report-out looks like at scale.
Themed session 11:00–11:45 The company clinical or commercial team on either the SIRO localised precision-medicine model or the AI reporting stack, chosen in advance because it determines the speaker.
Open questions and closing discussion 11:45–12:00 The programme lead and the session speakers remain for questions on transfer scope, registration and what the company cannot do.
Field stations and roof garden By arrangement, 12:00–12:30 The multi-graft tomato tree, perennial rice and the roof garden. Included for agriculture and food agendas or a lighter close; dropped when the delegation wants the extra session time.

The cost of reading a genome and the number of people who can be reached with it are two lines on one chart, and the whole business is a bet on how far apart they still are.

Two variations change this shape. A hospital delegation usually trades the exhibition time for a longer session on SIRO. A research delegation usually trades the field stations for more time on the multi-omics platform. Both are settled during planning rather than on the day.

Where a BGI Genomics visit can be arranged

BGI Genomics sits in Yantian on the eastern edge of Shenzhen, and that is where the programme above runs. It works as the life-sciences anchor inside a Shenzhen tech tour, which is a useful counterweight in a city whose technology itineraries otherwise run to hardware, batteries and robotics.

Most delegations pair it with one or two other stops. A Mindray medical visit is the natural companion for a healthcare delegation, because it shows the other half of Chinese medical technology — devices rather than diagnostics — and the two companies make a clean comparison on how China's healthcare exporters go abroad. Because the site is a full hour from the Nanshan and Futian clusters, put it first or last in the day rather than in the middle.

Practical notes for the day

Decide the theme before you book. The session speaker comes from the team that owns the subject, so the localised precision-medicine route and the AI reporting route are staffed by different people. Asking to cover both in full produces a rushed hour.

Bring someone who reads reports. The instruments and the galleries are legible to any executive; the part that repays attention — variant classification, interpretation, counselling workload, what the software does and does not decide — needs a clinician or a laboratory manager in the room. A delegation composed entirely of general managers will spend the session watching a demonstration.

Plan for interpretation and for terminology. Genomics vocabulary does not translate cleanly, and the terms that carry the meaning are the ones with no common equivalent: variant classification, carrier screening, minimal residual disease. Professional English interpretation is part of how we run the programme rather than an optional extra, and a glossary goes out with the pre-visit material.

Treat the distance as part of the programme. Dameisha is roughly an hour from central Shenzhen in normal traffic and longer at peak, the basement canteen has been open to organised groups, and the coastal path is a few minutes from the gate. Delegations that try to combine this half-day with two more visits in the same day consistently report that the value drops.

What previous delegations asked us to change

Three adjustments come up repeatedly when we review a programme with the groups that have just finished it, and all three are easier to decide before the booking than after.

Less museum, more session. The exhibition floors are excellent and they are not the reason a professional delegation came. Groups consistently ask to compress the gallery time and extend the themed session, and where the host's schedule allows we now build the morning that way.

A narrower session theme. A request for an overview of Chinese genomics produces an overview, which is what the galleries already gave. Groups that specify one question — how in-hospital capability is procured, what a transfer project contains, how AI is validated for clinical reporting — get a materially better hour, and it has to be requested in advance because it determines the speaker.

One stop, not three. Delegations that schedule this visit between two others in central Shenzhen lose two hours to the road. Groups that give it the whole morning and put a second company after lunch report a better day.

Questions before a BGI Genomics visit

Where is the visit held? At the BGI Time-Space Centre in Dameisha, Yantian District, Shenzhen, about an hour from Futian or Nanshan. The exhibition floors, the sequencing hall viewing corridor and the meeting rooms are all on the same site.

Can we go inside the sequencing laboratory? No, not as a matter of course. The fifth-floor corridor looks into the laboratory through glass, and that is the standard route. Entry into the laboratory itself is at the host's discretion on the day and should not be built into the agenda.

Is this a factory tour? No. This is a headquarters, exhibition and laboratory-building visit. The instruments on the other side of the glass are made by MGI Tech, a sister company under the same group, and instrument manufacturing is not part of this programme.

Who should be in the room? A clinical or laboratory lead alongside the commercial decision-makers. The session assumes a working knowledge of testing workflows, and delegations that bring only general managers get less from it.

Can we see the AI platforms? Yes, in the themed session. iGeneT Pro for analysis, variant filtering and report drafting, ChatGeneT for genetic counselling, both inside the BGI Zhihui medical system, and i99 for ongoing health management. The variant knowledge bases and the Shenzhen Data Exchange listing are covered in the same session.

Can the session topic be tailored to our team? Yes. The standard themes are the SIRO localised precision-medicine model and the AI reporting stack, and a third option — a roundtable on screening economics for public-health delegations — is available by arrangement. The theme has to be confirmed before the visit because it determines the speaker.

Is English interpretation provided? Yes. English interpretation on site is part of how we run the programme, and it matters most in the session, where clinical terminology carries the meaning. A glossary goes out with the pre-visit material.

What are the photography and security rules? Photography is restricted in the laboratory areas, and the corridor is a viewing route rather than an access route. Foreign visitors need passport details registered with the site in advance, which is why the lead time is what it is.

How far in advance should we book? Around 30 days is the practical lead time. The site takes organised groups rather than walk-ins, runs one visiting group at a time, and the speakers are drawn from active clinical and commercial teams, so the date is the binding constraint rather than the group size.

What is the minimum group size? The programme starts at 10 participants. Larger delegations are workable and we plan the room and the route split around the number you bring.

Arranging your BGI Genomics visit

We arrange the visit for you end to end: the introduction to the host, the scheduling on the Chinese side, the session theme, visitor registration, transport from your hotel and back, and English interpretation on the day. Tell us your dates, your delegation size and which theme your team needs, and we will confirm what can be arranged. Start at chinatechvisits.com/company-visits.