Article

Winheberg network latency from around the world (2026)

Real latency between Winheberg's Lyon infrastructure and over 150 cities worldwide, measured with mtr.tools on 24 June 2026, with detailed tables by region.

July 21, 2026 Winheberg
HomeBlogWinheberg network latency from around the world (2026)
Back to blog

At Winheberg, we talk about performance a lot, but we would rather let the numbers speak for themselves. We used mtr.tools, an independent service that runs ping probes from many measurement points spread across the world, to assess the latency between our infrastructure and the rest of the planet.

The results were collected on 24 June 2026. Each probe sent 10 packets to our network, and the data below reflects the average latency measured at that moment. These are neither guaranteed values nor contractual minimums, but a real snapshot of what our connectivity delivers from various regions of the globe.

What is latency?

Latency, or ping, is the time a data packet takes to make the round trip between point A and point B. It is measured in milliseconds (ms). The lower the value, the more responsive the communication.

For a game server, latency has a direct impact on the player experience. Below 30 ms, the game feels smooth and actions are almost instant. Between 30 and 50 ms, the experience stays comfortable for the vast majority of titles. Beyond 50 ms, players start to notice a lag that makes the game unpleasant, especially on real-time titles like FiveM or competitive FPS games.

For a VPS hosting a website or an application, latency affects the response time perceived by visitors. A site served from Lyon to a visitor in Paris will respond in a few milliseconds, whereas a visitor in Tokyo will wait more than 200 ms for the network trip alone.

How to read the tables

Each table shows the probe city, the operator hosting it and the average latency measured over the 10 packets sent. Results are sorted by increasing latency in each section. Keep in mind that these values depend on the network path taken between the probe and our infrastructure, not only on geographic distance. Two probes in the same city can show different results depending on their operator and its peering agreements.

France

Our infrastructure is hosted in Lyon. The local Winheberg probe logically shows 0.25 ms, the latency of the internal network. From Paris, we find around 7 ms, matching the propagation time of the optical fibre between Lyon and Paris. Roubaix, in the north, shows 10 ms. Strasbourg rises to 23 ms, likely because of routing through an external exchange point rather than a direct path.

City Operator Average latency
Lyon Winheberg 0.25 ms
Paris OneProvider 7.32 ms
Roubaix Smishcraft 10.05 ms
Strasbourg RackNerd 23.17 ms

Benelux

The Netherlands concentrates a large share of European exchange points, notably AMS-IX in Amsterdam. The Dutch probes show between 14 and 19 ms, and Belgium sits in the same range at 16 ms. These results reflect the peering density between French and Dutch networks.

City Operator Average latency
Amsterdam, Netherlands Privex 14.41 ms
Amsterdam, Netherlands ParadoxNetworks 14.86 ms
Amsterdam, Netherlands iFog 15.15 ms
Zoetermeer, Netherlands Serverion 15.54 ms
Amsterdam, Netherlands Scaleway 16.20 ms
Dessel, Belgium EDPNet 16.50 ms
Doetinchem, Netherlands Hybula 16.94 ms
Eygelshoven, Netherlands Atomic Networks 17.44 ms
Amsterdam, Netherlands xTom 18.23 ms

United Kingdom and Ireland

The United Kingdom is linked to the continent by several submarine cables crossing the Channel. From London, the probes show between 12 and 22 ms depending on the operator. Cambridge and Coventry sit in the same range. Dublin in Ireland, reached via the cables connecting France and the United Kingdom to Ireland, shows 23 ms.

These values are perfectly suited to gaming. A London player on a FiveM server hosted with Winheberg will notice no difference from a local server.

City Operator Average latency
London OneProvider 12.07 ms
London Mythic Beasts 12.79 ms
London Infraly 14.25 ms
Cambridge Mythic Beasts 14.33 ms
London HostHatch 16.03 ms
Coventry HostUS 16.11 ms
London Scaleblade 16.78 ms
London BerryByte 17.17 ms
Newcastle Innova Communications 19.11 ms
London ParadoxNetworks 20.61 ms
Peterborough MyHostingSpace 20.87 ms
London FREAKHOSTING 21.73 ms
Cambridge Fast2host 22.43 ms
Dublin, Ireland OneProvider 23.24 ms

Germany, Switzerland and Austria

Frankfurt is one of the largest network interconnection hubs in Europe, with DE-CIX. Local probes show between 15 and 24 ms, which reflects the strong peering between French and German networks. Nuremberg, a little further east, hovers around 17 to 18 ms.

Switzerland sits between 20 and 24 ms from Zurich. Vienna in Austria comes in between 23 and 25 ms. These values remain well below the threshold where a player or a visitor would notice any delay.

City Operator Average latency
Frankfurt OneProvider 15.13 ms
Frankfurt Datapacket 15.19 ms
Frankfurt level66.network 15.33 ms
Nuremberg Advin Servers 17.08 ms
Nuremberg netcup 18.35 ms
Frankfurt xTom 19.68 ms
Zurich, Switzerland OneProvider 19.88 ms
Frankfurt Netrouting 20.56 ms
Frankfurt meerfarbig 21.75 ms
Frankfurt ParadoxNetworks 22.13 ms
Vienna, Austria HostHatch 22.98 ms
Dusseldorf OneProvider 23.00 ms
Zurich, Switzerland Datapacket 23.60 ms
Zurich, Switzerland HostHatch 23.73 ms
Frankfurt iFog 23.93 ms
Vienna, Austria Alwyzon 24.88 ms

Iberian Peninsula and Italy

Barcelona shows 11 ms, close to the French cities thanks to geographic proximity and strong Mediterranean peering. Madrid hovers around 22 ms, and Lisbon at 48 ms. Routing to Portugal often goes through Madrid or the Atlantic cables, which adds a few milliseconds.

Florence sits at 33 ms and Palermo, at the very south of Italy, at 42 ms. Crossing the Alps and the length of the Italian peninsula explain this gap.

City Operator Average latency
Barcelona, Spain OneProvider 10.92 ms
Barcelona, Spain iFog 11.29 ms
Madrid, Spain OneProvider 21.86 ms
Madrid, Spain GINERNET 21.89 ms
Florence, Italy OneProvider 33.19 ms
Palermo, Italy Melbicom 41.59 ms
Lisbon, Portugal OneProvider 47.75 ms

Northern Europe

The Nordic countries are linked to the rest of Europe by submarine and land cables passing through Germany and Denmark. Stockholm and Oslo sit between 31 and 39 ms, consistent with the distance. Copenhagen shows 34 ms. Helsinki, further east, rises to 57 ms.

Reykjavík in Iceland sits at 50 ms. The DANICE submarine cable linking Iceland to Denmark explains this result, which stays reasonable given the island's geographic isolation.

For gaming, Stockholm and Oslo stay below 40 ms, which is comfortable. Helsinki at 57 ms starts to reach the limit.

City Operator Average latency
Stockholm, Sweden Privex 30.73 ms
Stockholm, Sweden OneProvider 31.04 ms
Stockholm, Sweden ParadoxNetworks 31.34 ms
Stockholm, Sweden Swehosting 31.47 ms
Oslo, Norway OneProvider 33.22 ms
Sandefjord, Norway WebHorizon 33.46 ms
Copenhagen, Denmark OneProvider 33.69 ms
Oslo, Norway iFog 34.17 ms
Stockholm, Sweden HostHatch 34.90 ms
Oslo, Norway SpanningNetworks 37.12 ms
Oslo, Norway HostHatch 39.37 ms
Reykjavík, Iceland Spaugur 49.50 ms
Helsinki, Finland Nicklas Y 56.83 ms

Eastern and South-Eastern Europe

Results vary more in this area. Poland hovers around 31 to 35 ms thanks to a strong land network through Germany. The Baltic states (Lithuania, Latvia, Estonia) sit between 43 and 49 ms, Romania and Bulgaria between 42 and 62 ms.

Turkey and Greece show latencies between 49 and 63 ms. Routing often goes through the Balkans or the Mediterranean submarine cables, which adds a few extra hops. Moscow comes in at 69 ms.

An interesting point is the variability between operators in the same city. In Bucharest for example, Datapacket shows 46 ms while HAZI.ro rises to 62 ms. The peering quality of each operator matters as much as distance.

City Operator Average latency
Prague, Czech Rep. OneProvider 29.33 ms
Warsaw, Poland Melbicom 30.74 ms
Prague, Czech Rep. Datapacket 30.98 ms
Warsaw, Poland Datapacket 31.98 ms
Wyszków, Poland FREAKHOSTING 33.29 ms
Warsaw, Poland Scaleway 34.36 ms
Belgrade, Serbia OneProvider 39.85 ms
Tirana, Albania OneProvider 40.96 ms
Sofia, Bulgaria Melbicom 41.60 ms
Kyiv, Ukraine OneProvider 41.68 ms
Vilnius, Lithuania OneProvider 42.85 ms
Riga, Latvia Melbicom 44.88 ms
Bucharest, Romania Datapacket 45.75 ms
Sofia, Bulgaria Talido 47.20 ms
Istanbul, Turkey Datapacket 48.78 ms
Tallinn, Estonia xTom 48.90 ms
Nicosia, Cyprus OneProvider 56.15 ms
Athens, Greece Fusioned 58.42 ms
Bucharest, Romania HAZI.ro 61.75 ms
Moscow, Russia Melbicom 68.97 ms

North Africa and the Middle East

The Middle East and North Africa are connected to Europe via the Mediterranean submarine cables, notably those crossing the Suez Canal. Fez in Morocco shows 39 ms thanks to the cables linking the Maghreb directly to France. Tel Aviv sits at 70 ms, matching the crossing of the eastern Mediterranean. Dubai rises to 108 ms and Fujairah to 123 ms, with routing that often goes through Northern Europe or India before coming back down to the Gulf.

City Operator Average latency
Fez, Morocco OneProvider 39.39 ms
Tel Aviv, Israel OneProvider 69.58 ms
Dubai, United Arab Emirates ServerWala 108.34 ms
Fujairah, United Arab Emirates Melbicom 122.51 ms

North America (East Coast)

Crossing the Atlantic imposes a physical floor around 75 to 80 ms between France and the East Coast of the United States. This is the time light takes to travel the thousands of kilometres of submarine optical fibre linking Europe to America. The New York probes show between 78 and 81 ms, consistent with this reality.

Montreal and Toronto in Canada sit around 91 ms. Chicago, further inland, rises to 96 to 100 ms. Beyond 50 ms, the gaming experience degrades noticeably, so the East Coast is not suited to gaming from our servers, but these values remain fine for web or application hosting.

City Operator Average latency
New York SMC Server Solutions 78.56 ms
New York GalaxyGate 78.75 ms
Secaucus, NJ Advin Servers 78.83 ms
New York HostHatch 79.05 ms
Piscataway, NJ SYSTEM36 79.15 ms
New York OneProvider 79.20 ms
Allentown, PA Crunchbits 79.54 ms
New York ExtraVM 80.67 ms
Ashburn, VA OneProvider 85.02 ms
Ashburn, VA ParadoxNetworks 85.02 ms
Ashburn, VA iFog 85.99 ms
Reston, VA OneProvider 87.10 ms
Montreal, Canada OneProvider 90.88 ms
Toronto, Canada OneProvider 91.83 ms
Chicago OneProvider 95.81 ms
Chicago HostHatch 96.67 ms
Ottawa, Canada EBOX 101.32 ms

North America (Central and South)

The central and southern United States sit between 110 and 128 ms. Miami hovers around 110 to 120 ms, Dallas and Houston between 121 and 127 ms. These values are explained by crossing the Atlantic added to the land trip inside the continent.

City Operator Average latency
Miami Advin Servers 110.46 ms
Miami ExtraVM 110.70 ms
Miami Datapacket 113.70 ms
Miami FREAKHOSTING 114.36 ms
Miami OneProvider 116.40 ms
Kansas City Macarne 116.76 ms
Monticello, IA iWebFusion 117.66 ms
Kansas City OneProvider 117.83 ms
Miami Netrouting 120.21 ms
Houston iFog 121.16 ms
Denver Datapacket 121.21 ms
Salt Lake City OneProvider 121.56 ms
Dallas OneProvider 123.18 ms
Dallas FREAKHOSTING 126.70 ms
Atlanta OneProvider 127.69 ms

North America (West Coast)

The West Coast combines crossing the Atlantic and the land trip from east to west across the North American continent. Los Angeles sits between 134 and 153 ms, San Jose and Fremont between 142 and 152 ms, Seattle around 138 ms.

These values are not suited to gaming, but remain perfectly fine for web hosting, APIs, or any use where a few tenths of a second of delay are not a problem.

City Operator Average latency
Los Angeles Hivelocity 133.93 ms
Mexico City, Mexico OneProvider 134.10 ms
Phoenix GTHost 134.95 ms
Los Angeles iWebFusion 135.00 ms
Los Angeles HostUS 135.02 ms
Bend, OR Tier.Net 135.11 ms
Los Angeles RamNode 137.68 ms
Seattle Pudu 137.89 ms
Los Angeles OneProvider 143.56 ms
San Jose DMIT 144.99 ms
Vancouver, Canada Free Range Cloud 145.46 ms
San Jose xTom 146.08 ms
Los Angeles Advin Servers 148.55 ms
Fremont Mythic Beasts 152.45 ms
Portland OneProvider 153.17 ms

South and Central America

Traffic between Europe and South America generally uses the Atlantic submarine cables linking Portugal or France to Brazil. From São Paulo, latency hovers around 195 to 204 ms. Costa Rica shows 167 ms, Bogota 178 ms, Lima 197 ms.

The Southern Cone countries (Argentina, Chile, Bolivia) rise beyond 230 ms. Routing often goes through São Paulo or the East Coast of the United States before coming back down, which lengthens the trip.

City Operator Average latency
San Juan, Puerto Rico OneProvider 158.42 ms
San Jose, Costa Rica OneProvider 166.66 ms
Bogota, Colombia OneProvider 178.23 ms
Quito, Ecuador OneProvider 190.90 ms
Sao Paulo, Brazil OneProvider 194.90 ms
Lima, Peru OneProvider 197.14 ms
Sao Paulo, Brazil Datapacket 204.11 ms
Valdivia, Chile ZappieHost 229.49 ms
Buenos Aires, Argentina OneProvider 240.17 ms
Santiago, Chile OneProvider 244.08 ms
La Paz, Bolivia OneProvider 251.33 ms

Asia

Asia is linked to Europe by two major submarine cable routes. The first goes through the Mediterranean, the Suez Canal and the Indian Ocean. The second takes the long way around through the Atlantic and the Pacific. Depending on the operator's peering, the path taken can vary considerably.

Mumbai in India shows 110 ms via the first route (Suez), a result consistent with the distance. Singapore sits between 150 and 156 ms, and South-East Asia (Malaysia, Thailand, Vietnam) stays in this 154 to 205 ms range.

East Asia (Hong Kong, Tokyo, Seoul) rises between 174 and 252 ms. The variability is significant depending on the operator. In Seoul for example, LG DACOM shows 205 ms while Korea Telecom rises to 247 ms.

City Operator Average latency
Mumbai, India OneProvider 110.51 ms
Mumbai, India ServerWala 142.56 ms
Hyderabad, India soundar 147.26 ms
Singapore HostHatch 149.54 ms
Singapore iFog 153.60 ms
Johor, Malaysia Advin Servers 154.13 ms
Singapore ExtraVM 155.62 ms
Bangalore, India ServerWala 156.66 ms
Kuala Lumpur, Malaysia OneProvider 158.91 ms
Delhi, India ServerWala 168.08 ms
Hong Kong MLNL.HOST 174.26 ms
Hong Kong HostHatch 179.88 ms
Hong Kong xTom 185.27 ms
Bangkok, Thailand OneProvider 188.25 ms
Taipei, Taiwan Fast Line 194.69 ms
Taichung, Taiwan Veetime 202.82 ms
Seoul, South Korea LG DACOM 205.45 ms
Hanoi, Vietnam VNPT 205.42 ms
Seoul, South Korea LG POWERCOMM 216.68 ms
Tokyo, Japan xTom 220.68 ms
Tokyo, Japan HostHatch 220.98 ms
Islamabad, Pakistan OneProvider 229.99 ms
Tokyo, Japan OneQode 230.64 ms
Tokyo, Japan DMIT 232.13 ms
Seoul, South Korea SK Broadband 235.00 ms
Seoul, South Korea Korea Telecom 246.94 ms

Sub-Saharan Africa and Oceania

Africa shows varied results. Lagos in Nigeria shows 115 ms, a result linked to the West African submarine cables (such as ACE and MainOne) that connect West Africa directly to France. Cape Town sits at 161 ms, Johannesburg between 163 and 180 ms depending on the operator, and Nairobi in Kenya at 181 ms.

Australia and New Zealand are among the regions furthest from France in terms of network routing. Sydney sits between 240 and 245 ms, Brisbane at 257 ms and Auckland at 274 ms. Traffic generally uses the submarine cables passing through South-East Asia or the Pacific.

City Operator Average latency
Lagos, Nigeria Melbicom 114.57 ms
Cape Town, South Africa Rackzar 161.06 ms
Johannesburg, South Africa Linux Tech 163.00 ms
Johannesburg, South Africa ZappieHost 163.63 ms
Durban, South Africa Linux Tech 164.17 ms
Johannesburg, South Africa Rackzar 180.21 ms
Nairobi, Kenya Wofbit Networks 180.53 ms
Sydney, Australia SMC Server Solutions 240.34 ms
Sydney, Australia xTom 242.35 ms
Sydney, Australia OneProvider 243.57 ms
Sydney, Australia Datapacket 245.33 ms
Brisbane, Australia RansomIT 257.28 ms
Auckland, New Zealand ZappieHost 274.33 ms
Guam OneQode 282.44 ms

Key takeaways

These measurements confirm what geography and physics led us to expect. With Winheberg servers hosted in Lyon, latency to Paris hovers around 7 ms, the travel time of the Lyon-Paris fibre. All of Europe stays below 60 ms. The East Coast of the United States sits around 79 to 86 ms, matching the physical floor of crossing the Atlantic. Beyond that, values rise proportionally to distance and depend on the routing taken by each operator.

The most decisive factor is not always geographic distance. Peering between operators, the path taken by submarine cables, and the quality of local interconnections play an equally important role. This is what explains why a probe in Barcelona can show 11 ms while another in Strasbourg, which is in France, rises to 23 ms.

You can reproduce this test yourself by going to mtr.tools and running a ping toward the IP of your Winheberg server.

Test your own latency

Winheberg provides a Looking Glass available at any time, which lets you run a ping or a traceroute from our infrastructure toward any destination.

It is available at https://lg.winheberg.com

Our servers are hosted in Lyon, on our own infrastructure. These results reflect the real latency measured from each region of the world 🌍

latencynetworkpingmtrperformanceinfrastructure