When connectivity is discussed in policy terms, it is usually the availability of infrastructure that gets framed as the challenge — the towers, the cables, the spectrum, the fibre runs, the question of whether a signal exists in a given valley at all. From inside Kyrgyzstan, in 2024, that framing has stopped describing the actual situation. Mobile coverage is, in most of the country, no longer the limit. The limit is what gets done with the connection that already exists.
One observation, from a road in southern Kyrgyzstan, makes the point more directly than the aggregate figures do. On the route between Tash-Kumyr and Kerben, in Jalal-Abad oblast, in October 2024, a few stretches of road returned download speeds of around 120 Mbps on the MEGA network when measured against both fast.com and Speedtest by Ookla. This was not part of a structured testing campaign. It was a one-off check during travel — done out of curiosity rather than methodology. For a remote rural road in a high-mountain southern region, in that month, the result was incredibly good.
That kind of reading, in that kind of place, sits awkwardly with most of the assumptions the connectivity story usually carries. The country has built, by any reasonable regional measure, a strong mobile network. Coverage reaches almost every populated settlement. Pricing is among the lowest in the world. Speeds in rural areas are often faster than in Bishkek. And yet much of the work that needs a stable connection — administrative tasks, document handling, video calls between agencies, school and clinic systems — still depends on a fixed line that, in most of the country, isn’t there.
What the coverage map actually means
Mobile networks reach 99% of Kyrgyzstan’s 2,227 populated settlements. The three operators — MEGA (state-owned), Nur Telecom / O! (Visor Holding, Kazakhstan), and Sky Mobile / Beeline (VEON) — split the market roughly evenly, and between them they have closed almost the entire physical access gap. As of mid-2024, active mobile subscriptions sit above 7.8 million — more than the population, because many users carry two SIMs.
The 22 settlements still without coverage are scattered across high-altitude mountain valleys. They are not waiting for a tower. They are waiting for electricity. Without a stable power source, the equipment cannot run; without the equipment, no operator extends a signal. This is a practical limit of terrestrial deployment, not a gap that more telecom investment will close.
Faster in the south than in Bishkek
One of the more counter-intuitive patterns visible from the field is that mobile speeds in rural areas are often higher than in the capital. Repeated readings put 4G downloads in the range of 90–120 Mbps in rural areas against 30–60 Mbps in congested parts of Bishkek and central Osh. The reason is straightforward: where fewer devices share a tower, each device gets more of its capacity. In urban centres, the same infrastructure carries a much higher density of simultaneous users.
Read across enough of these moments, the rural digital divide in Kyrgyzstan looks less like an access gap and more like a usage gap. Signal exists. Speed is, in many cases, better than in the capital. What is uneven is what people do with the connection: which services are available in Kyrgyz or Russian and worth using, which institutions have moved their workflows online, and which households can afford devices and applications that match what the network can carry.
What 120 Mbps actually enables
The number on a speed test screen is easy to print and easy to ignore. It is more useful read against the kind of work it makes possible. At 120 Mbps download on a single device, on a stretch of road in Jalal-Abad, the connection comfortably handles services that, ten years ago, would have been considered urban-only or institutional:
- 4K video streaming — Netflix, YouTube and similar platforms recommend roughly 25 Mbps for a single 4K stream; at 120 Mbps a household could in principle run multiple simultaneous high-definition streams without contention.
- Multi-party video conferencing — a stable Zoom, Google Meet or Teams group call needs around 3–4 Mbps per HD participant; the connection has headroom for an entire team to meet from different rooms in the same building, or for a teacher to host a remote class.
- Telehealth and remote diagnostics — transmitting medical imaging or running a tele-consultation with a specialist in another oblast becomes technically feasible from a rural clinic that once would have had to send the patient to Bishkek.
- Telemetry and IoT — fields of low-bandwidth sensors (LoRaWAN gateways, weather stations, agricultural monitors, environmental sensors) can backhaul their data to a central platform without straining the link.
- Cloud workflows for SMEs — accountants, lawyers, designers and small administrative offices can work from cloud-hosted documents and tools at the same speed they would expect in a city office.
- Distance learning and large file transfers — students downloading course materials, software updates, or research datasets do not have to plan around the connection.
The list describes what the connection can carry, not what it currently carries. A 120 Mbps signal on a rural road has the technical capacity for telehealth; the platform, the licensed practitioners, the records system, and the patient referral pathway sit elsewhere on the institutional map. The same goes for most of the other items. Where the network has become strong enough, the next layer of constraints starts to show — and most of them are not radio constraints.
The fixed line that mostly isn’t there
Where mobile coverage is near-universal, fixed broadband at home is not. Household fixed broadband penetration sits at about 21%. The Asian regional average is roughly three times that. The terrain is part of the answer — laying fibre to villages strung along mountain roads is expensive — and the rest is economic. A 70–100 GB mobile data plan costs less than ten dollars a month. For most households that already cover their needs through a phone, there is little reason to also pay for a fixed line.
| Metric | Kyrgyzstan | Asia (regional avg.) |
|---|---|---|
| Household fixed broadband penetration | ~21% | ~63.7% |
| Mobile broadband as share of mobile subscriptions | 91% | — |
| Settlements with mobile coverage | 99% | — |
Where it matters most is not in households but in institutions. Schools, clinics, and district administrations need fixed connections that a single mobile SIM cannot replace — for record systems, for video, for any service that depends on consistent throughput across a working day. The 120 Mbps reading on a road in Jalal-Abad is impressive in the household sense, where one device meets one network. It does not, by itself, tell you anything about what happens when a district school or rural clinic tries to put forty users on the same connection at once.
Digital CASA and the institutional gap
The World Bank’s Digital CASA project is the largest current effort to address the fixed-line gap. It plans to lay 2,500 km of fibre-optic cable and connect roughly 4,000 government facilities — hospitals, schools, administrative offices — to a high-speed network. As of late 2024, about 15% of the planned cable was in the ground. When complete, the project is expected to put around 60% of the population within reach of fixed broadband, indirectly through the institutional anchor sites.
Read against the field, the picture rearranges itself. Mobile, at the household level, is a story that has largely worked. Fixed broadband, where it matters, is an institutional question — about which schools can run a digital lesson without dropouts, which clinics can transmit imaging, which oblast administrations can share documents in real time. Cable kilometres are easy to count. The harder count is how many of those buildings are actually wired in once the fibre passes the door.
What the signal does and does not solve
The 120 Mbps reading on a roadside in Jalal-Abad was not, in the most ordinary way, a sign that everything works. It was a small reminder of what infrastructure availability now does and does not settle. Almost every populated settlement in Kyrgyzstan has a mobile signal. Most households have mobile internet. Speeds in much of the country are perfectly adequate — and in some places more than adequate — for the everyday tasks people use the network for. None of that, on its own, builds a digital state. The next set of questions has moved off the radio.
The coverage map is largely complete. The institutional map is not.
The next two pieces in this series move outward from the household — to the backbone, the cables and exchanges and transit routes that carry the country’s traffic in and out, and to affordability, a quiet pricing advantage Kyrgyzstan has held without making much of it.
The Digital Infrastructure assessment for the Kyrgyz Republic was prepared by Aziz Soltobaev as part of the DECA assessment. The observations and field readings cited here are drawn from KG Labs’ own intelligence and field research. For the underlying data, methodology, or a tailored briefing for your programme, get in touch.