Travel eSIM Latency: What Really Determines Speed?
Travel eSIMs are sold like simple digital travel products: choose a country, pick a data allowance, install, and land connected. What the checkout page rarely shows is the telecom supply chain underneath.
New Q2 2026 Speedtest data maps more than 50 travel eSIM brands to the wholesale networks behind their profiles, the countries where their traffic reaches the public internet, and the performance consequences. Two travelers can stand on the same street, use the same local mobile network and still experience dramatically different latency because their data takes very different journeys after leaving the radio network.
Travel eSIM comparison has spent years obsessing over gigabytes, coverage and price. The data suggests the next battleground is less visible: routing architecture.
The brand is only the front door
A travel eSIM has a retail identity and an infrastructure identity. The consumer may see Airalo, Holafly, Ubigi or Nomad eSIM, but the underlying SIM profile is normally anchored to a separate wholesale or mobile network.
The Speedtest analysis found a concentrated infrastructure layer. KPN is Airalo’s largest measured wholesale anchor, while its profiles also appeared on Webbing, Orange, Play and Telna. Holafly’s samples were mainly linked to Hutchison Austria and China Mobile Hong Kong. Singtel supported brands including Nomad and Eskimo.Travel; Transatel supported Ubigi; Proximus and BICS appeared behind Roamless and Firsty; and Sky UK anchored the eSIM Go platform.
READ MORE: Everyone’s Selling eSIMs—But Who Actually Runs the Network?
That means different retail brands can share infrastructure, while one brand can use several wholesale partners depending on destination or product. Judging an eSIM company as if it were one fixed network is increasingly misleading.
| What travelers see | What may actually determine performance |
|---|---|
| “5G included” | Whether 5G access exists in the wholesale agreement |
| Local partner network | Where traffic exits to the public internet |
| “Unlimited data” | Fair-use rules or deliberate speed caps |
| Global coverage | Number and location of regional exit points |
| One retail brand | Several suppliers and profile types |
The exit point changes everything
The most important finding is not which provider posted the highest download number. It is where the data goes.
In home-routed roaming, traffic is tunneled from the visited network toward the home or wholesale core before reaching the public internet. GSMA specifications support both home-routed and local-breakout architectures; local breakout lets traffic access the internet through the visited network instead.
READ MORE: Why Two eSIMs on One Network Feel Different?
Travel eSIMs still overwhelmingly favor home-style routing. The Speedtest dataset identifies single-hub designs, distributed cores with several regional exits, and comparatively rare local breakout. In Japan, local au users recorded 36 ms median latency. A locally routed travel eSIM profile reached 54 ms, while a Madrid-routed profile on the same underlying radio network reached 543 ms.
That is the part the 5G icon cannot tell you. A phone can have an excellent radio connection and still feel sluggish because every interaction is taking a long detour. For messaging or background downloads, that may be acceptable. For video calls, cloud tools, gaming and remote desktop work, it can be very noticeable.
Distributed cores are becoming an advantage
Saily was measured using exits in Germany, the Netherlands, the United States, Singapore and Japan and recorded a global median multi-server latency of 117 ms. 1GLOBAL, with US and European hubs, recorded 121 ms. Webbing kept nearly three-quarters of measured traffic in the same country as the test. Airalo sat between single-hub and distributed models because its multi-supplier structure produced exits in several regions.
Long-distance hub routing produced brutal penalties: Amsterdam-routed profiles in Australia reached 30 times local latency; Madrid-routed profiles in Singapore reached 25 times local latency; and Singapore-routed profiles in the United States reached 17 times local latency.
READ MORE: The Latency Report: Ranking Global eSIMs by Jitter and Packet Loss for Remote Professionals
This fits closely with Alertify’s own Latency Report, where we have argued that “fast eSIM” is a poor category unless routing, jitter, packet loss and route consistency are considered alongside raw speed. Earlier Alertify comparisons also put Ubigi among the stronger performers for latency and infrastructure credibility.
There is now stronger evidence for why: architecture, not branding, sets much of the performance ceiling.
Speed is also a commercial decision
Throughput can reflect tariff design as much as network quality.
Yesim recorded a global median download speed of 6.03 Mbps and Instabridge 10.23 Mbps, patterns the source interprets as consistent with throttled unlimited plans. Firsty’s free tier sat at 0.21 Mbps. At the higher end, 1GLOBAL and Nomad were around 80 Mbps, Ubigi 75.89 Mbps and Holafly 56.87 Mbps.
The useful distinction is that speed and latency are not interchangeable. Holafly is a good example: respectable bulk download throughput can coexist with the latency penalty of a distant hub. A lower-speed plan with a nearby exit may feel more responsive in everyday use.
READ MORE: Study Exposes How Travel eSIMs Reroute Data Through China — and Let Anyone Become a Reseller
That makes simplistic “fastest eSIM” rankings questionable. A business traveler on Teams has different needs from someone downloading video at a hotel, and both differ from a tourist who mainly needs maps, messaging and ride-hailing.
The same applies to 5G. The source found major brands ranging from roughly one-fifth to more than 60% 5G usage in measured samples, with Ubigi leading at 62%, suggesting that wholesale access rights shape how much 5G a travel eSIM actually receives.
The supply chain keeps moving
Wholesale relationships are changing quickly. During the year to June 2026, Airalo increased the share of measured profiles anchored on KPN, reduced Orange France exposure and added Telna. Holafly consolidated further around Hutchison Austria. Saily moved from rented wholesale anchors toward its own registered network and, by late 2025, was carrying 95% of measured samples through its own distributed core.
Owning more of the core gives a provider greater control over routing, egress geography and optimization. It does not guarantee the best experience everywhere, but it reduces dependence on another supplier’s architecture.
READ MORE: Travel eSIMs vs Local SIM Cards: Real Prices, Real Scenarios
Academic research points in the same direction from another angle. A 2025 USENIX Security study examined 25 travel eSIM providers and highlighted how third-party routing and limited infrastructure transparency can create privacy and jurisdiction questions. An ACM Internet Measurement Conference study of Airalo across 24 countries likewise found a multi-operator roaming ecosystem rather than a simple “one brand, one network” model.
For consumers, opacity is mainly a performance issue. For enterprises, it can also become a governance issue.
What travelers should compare
Most providers do not publish enough routing detail for a normal customer to make a perfect decision, and asking tourists to understand packet gateways before buying 10 GB would be absurd.
But comparisons can improve. They should increasingly ask: Which local networks can the plan use? Does the provider use one global hub or several regional exits? Is local breakout available? Are speeds capped? Is 5G broadly available? Does the provider control its own core?
Local SIMs and operator visitor eSIMs remain worth considering where maximum local performance matters. Distributed-core travel eSIMs are a practical middle ground for travelers who value easy installation and multi-country use. A single-hub product may still be perfectly adequate for maps, messaging and casual browsing, particularly when price or unlimited data matters more than responsiveness.
Conclusion
The travel eSIM market spent its first growth phase making roaming cheaper and easier to buy. Its next phase will be about making the invisible network underneath that purchase better.
That changes how providers should be judged. Airalo’s scale and multi-supplier model give it reach and flexibility. Ubigi benefits from Transatel’s established telecom infrastructure. Saily’s move toward its own distributed network gives it more architectural control. Holafly can deliver strong throughput and a compelling unlimited-data proposition, but the Speedtest findings show why a distant hub can remain a weakness for latency-sensitive users. None of these models wins every trip.
The bigger shift is that wholesale routing is becoming measurable. Once customers, enterprises and comparison platforms can see that a 5G travel eSIM may be sending traffic halfway around the world before opening a webpage, “200+ countries” and “unlimited data” stop being enough.
Price per gigabyte built the category. Routing quality, recovery, transparency and infrastructure control are likely to define its more mature version. For Alertify, that is where eSIM comparison needs to go next: away from storefront claims and deeper into the network that actually delivers the experience.
