BytePlus CDN is not obviously expensive. It is expensive to misunderstand.
The first thing most teams see is the headline traffic rate. In Europe and North America, BytePlus starts at $0.044/GB for the first 10 TB each month, which is reasonable enough to put BytePlus CDN on the shortlist. But the invoice does not stop at traffic. BytePlus also charges for requests, origin transfer, log delivery, and billable 4xx responses. Regional pricing jumps sharply in some markets, and the free trial is more nuanced than it first appears because the trial resource packages and savings plans follow different validity periods.
That is the core tradeoff. BytePlus CDN can make sense for engineering-led teams that want a more platform-style CDN with logs, APIs, and room for edge-heavy workflows. But teams that want a clean monthly forecast should model the full billing stack before assuming the headline per-GB rate tells the story.
Quick answer: is BytePlus CDN expensive or not?
Short answer: BytePlus CDN is not expensive on traffic alone in lower-cost regions, but it can become expensive in practice if your workload generates lots of requests, frequent origin fetches, paid logs, or abusive 4xx traffic.
For Europe and North America, the first 10 TB each month starts at $0.044/GB. That is competitive enough for serious evaluation. The problem is that BytePlus does not behave like a single-line-item CDN. Traffic is only one part of the bill. You also need to account for $0.012 per 10,000 requests, $0.06/GB for origin transfer, paid logging, and region-specific rate differences that can materially change the final number.
If your traffic is concentrated in Europe or North America, your cache hit ratio is high, and your team is disciplined about monitoring and abuse control, BytePlus can be commercially reasonable. If your workload is request-heavy, log-heavy, or spread into more expensive regions, the service becomes harder to forecast and easier to underestimate.
What BytePlus CDN charges for
BytePlus documents five main CDN billing items under pay-as-you-go pricing:
| Billing item | How BytePlus bills it | Published rate |
|---|---|---|
| Traffic from CDN to client | By traffic, region-specific, tiered by monthly usage | Varies by region and tier |
| Origin transfer | By traffic | $0.06/GB |
| Requests received by CDN | By request count | $0.012 per 10,000 requests |
| Log Batch Delivery | By log entries delivered | $0.001 per 10,000 entries |
| Log Streaming | By log entries delivered | $0.0015 per 10,000 entries |
Two mechanics matter here.
First, BytePlus settles CDN traffic hourly, but the traffic tier depends on your cumulative monthly usage in each billing region. That is more complex than a flat global card because the effective rate changes as volume accumulates.
Second, BytePlus is not request-inclusive. That matters for small files, APIs, HLS or DASH segment delivery, image-heavy sites, and anything else where request count climbs faster than transferred GB.
BytePlus regional traffic pricing
The traffic card is where BytePlus looks straightforward at first glance and more complex on closer inspection. For the first 10 TB each month, the published rates include:
| Region | First 10 TB per month |
|---|---|
| Europe | $0.044/GB |
| North America | $0.044/GB |
| Middle East/Africa | $0.100/GB |
| South America | $0.100/GB |
That spread is significant. A workload that costs about $440 for 10 TB in Europe or North America would cost about $1,000 for the same 10 TB in Middle East/Africa or South America before requests, origin transfer, or logging are added.
BytePlus also uses descending volume tiers, so the unit rate falls as usage grows within a region. That helps at larger volumes, but it does not remove the need to model traffic geography carefully. If your audience mix shifts across regions, your effective blended rate can move more than expected.
For large or enterprise traffic volumes, BytePlus also says customized pricing is available. That matters because the public tables are detailed, but they are not necessarily the whole commercial story for negotiated buyers.
Request fees, origin fees, and log fees
The easiest way to misread BytePlus pricing is to focus only on the per-GB card and ignore the secondary line items.
Requests: BytePlus charges $0.012 per 10,000 requests. That looks small until request volume gets large. At 200 million requests per month, request charges alone add about $240. For request-dense workloads, that is not a rounding error.
Origin transfer: BytePlus charges $0.06/GB for data transferred from CDN to origin. This makes cache efficiency commercially important. A poor cache hit ratio does not just hurt performance. It creates another billable layer.
Logs: BytePlus charges $0.001 per 10,000 entries for Log Batch Delivery and $0.0015 per 10,000 entries for Log Streaming. Those headline fees are only part of the picture. The documentation also says:
- Log Batch Delivery adds object-storage fees, such as BytePlus TOS, AWS S3, or another S3-compatible destination.
- Log Streaming adds TLS fees for storage and log read/write traffic.
That means logs can create a two-part cost: a CDN-side per-entry fee plus downstream platform fees.
There is also a cost-control nuance here. BytePlus documents Log Download as free, but slower. If near-real-time log analysis is not operationally necessary, some teams can avoid the paid log paths by using the delayed download option instead.
A simple example shows how the stacking works. Assume a Europe workload with:
- 10 TB of client delivery
- 200 million requests
- 1 TB of origin transfer
- 200 million log entries through Log Batch Delivery
Using the published rates, that lands around:
- $440 for client traffic
- $240 for requests
- $60 for origin transfer
- $20 for batch log delivery
Total: about $760/month, before taxes, before storage charges for the destination bucket, and before any abuse-related overage. Replace Batch Delivery with Log Streaming and the CDN-side log line rises again, while TLS charges still sit outside the CDN rate card.
Hidden cost risks, including billable 4xx and abuse traffic
This is where BytePlus pricing becomes more operational than many teams expect.
4xx responses are billable. BytePlus explicitly says 4xx responses generate billable traffic, and those requests also count toward billable request volume. In plain terms, rejected, blocked, or broken requests can still cost money because the CDN still returns response headers and bodies.
That matters for bot traffic, scraping, hotlinking, misconfigured clients, and traffic-theft scenarios. BytePlus has a dedicated document on CDN traffic theft risk that warns abuse spikes can produce high bills and even lead to service suspension if bills go overdue.
There are three more buyer caveats in the docs:
- Bills can exceed raw log size. BytePlus says billed traffic can be higher than the traffic size shown in logs because billing includes the file itself, TCP/IP and Ethernet header overhead, and TCP retransmissions.
- Real-time monitoring itself can cost extra. BytePlus recommends Log Streaming to analyze suspicious traffic quickly, but the same documentation notes that real-time logs are a paid feature and also carry downstream TLS fees.
- Usage caps are useful, but not instant. BytePlus documents a Usage Cap feature that can disable a domain after a threshold breach, but it warns of about a 15-minute delay between crossing the threshold and actual disablement. Usage during that window is still billed.
There is also some operational friction in the billing model. BytePlus says alternative billing methods and service unsubscription require a manual ticket rather than a self-serve toggle. That is not a direct fee, but it does affect how quickly finance or operations teams can respond when costs move in the wrong direction.
Savings plans and trial caveats
BytePlus offers prepaid savings plans for traffic and requests, but teams should read the published tables carefully rather than assume the word “savings” always means a better unit price.
The pricing page says savings plans are valid for 30 days and can offset data transfer and request charges. It also lists standard public pricing for those plans. For example, the standard outside Chinese Mainland traffic plan is priced at:
- $60 for 1 TB
- $300 for 5 TB
- $600 for 10 TB
That equals $0.06/GB. For Europe and North America first-tier traffic, that is actually higher than the published pay-as-you-go rate of $0.044/GB. For more expensive regions such as South America or Middle East/Africa, the same plan can look better. The point is simple: the public savings-plan math is not uniformly cheaper across all geographies.
The request savings plan is even less obviously compelling on the public card. BytePlus lists $12 for 10,000,000 requests, which works out to the same unit rate as the pay-as-you-go request fee. BytePlus also says the actual price is what appears on the purchase page, so teams should confirm whether their real quote contains a meaningful discount.
The free-trial wording is easier to read once you separate the two benefit types:
- The free trial resource packages, covering 1 TB of traffic and 10,000,000 requests, are valid for 1 year from the claim date.
- The free-trial savings plans described on the pricing page are valid for 30 days.
In other words, teams should not treat the one-year package validity and the 30-day savings-plan validity as the same thing. They describe different parts of the offer.
When BytePlus pricing makes sense
BytePlus pricing makes the most sense when your team is comfortable managing a multi-variable CDN bill instead of expecting a simple traffic invoice.
It is a sensible fit when:
- Your team is engineering-led and comfortable modeling traffic, requests, cache hit ratio, and logging behavior.
- Your traffic is concentrated in lower-cost regions or large enough to justify negotiated pricing.
- You genuinely need platform-style operational features such as detailed logs, API workflows, or more advanced delivery controls.
- You can support active abuse monitoring and cost guardrails, not just passive pay-as-you-go usage.
BytePlus also looks more defensible for workloads where the broader platform matters, such as teams already leaning into BytePlus tooling, security add-ons, or edge-oriented delivery operations. In those cases, higher pricing complexity may be acceptable because the priority is platform capability, not delivery simplicity. For a broader product picture beyond pricing, see this BytePlus CDN review.
BytePlus vs CDNsun on published pricing
This is where the comparison becomes more concrete. On published pricing, CDNsun is not only simpler to model. In several important traffic regions, it is also cheaper on raw delivery rates.
| Region | BytePlus published traffic rate
|
CDNsun published traffic rate
|
Published gap |
|---|---|---|---|
| Europe | $0.044/GB | $0.030/GB | CDNsun lower by $0.014/GB |
| North America | $0.044/GB | $0.030/GB | CDNsun lower by $0.014/GB |
| South America | $0.100/GB | $0.060/GB | CDNsun lower by $0.040/GB |
| Africa | $0.100/GB | $0.060/GB | CDNsun lower by $0.040/GB |
The most practical example is Europe or North America at 10 TB per month. On traffic alone, BytePlus comes to $440. CDNsun comes to $300. That is a published gap of $140/month in CDNsun’s favor before BytePlus request fees, origin transfer fees, and log charges are added.
The cost-structure difference is just as important as the traffic-rate difference. BytePlus adds separate billing lines for requests, origin transfer, and paid logs. CDNsun’s published model is more bundled for common delivery workloads:
- No monthly fee
- Unlimited requests included
- Raw logs included
- No extra charges for streaming
- Integrated CDN storage, with free traffic from CDNsun storage to CDNsun edge
If your workload is closer to mainstream website or video delivery, CDNsun also documents raw access logs publicly. If you want to benchmark that simpler model directly, CDNsun offers a 15-day free trial and you can sign up here.
That does not prove CDNsun is cheaper in every workload, and this article is not making that claim. But where the public rate cards are directly comparable, CDNsun has a clear price advantage, and the surrounding billing model is easier to forecast because fewer secondary charges sit outside the traffic line. For the Middle East specifically, teams should verify the current CDNsun rate card separately rather than infer it from the Africa comparison above.
Final verdict
BytePlus CDN pricing is documented, but it is not simple.
The public traffic rates for Europe and North America are reasonable enough to attract attention. But the real risk sits in the surrounding mechanics: request charges, origin transfer, paid logs, regional price variation, billable 4xx responses, and abuse traffic that can still generate meaningful cost even when requests are being blocked or rejected.
If you want a platform-style CDN and your team is ready to manage pricing as an operational system, BytePlus can make sense. If you want a cleaner commercial model for mainstream content delivery, CDNsun is not only easier to forecast. In the published Europe, North America, South America, and Africa comparisons above, it is also cheaper.
Either way, do not approve BytePlus from the per-GB card alone. Model the full invoice path first.
FAQ
Does BytePlus charge for requests?
Yes. BytePlus charges $0.012 per 10,000 requests. That matters most for small-object workloads, image-heavy sites, API-adjacent delivery, and segmented video traffic where request count can rise faster than transferred GB.
Does BytePlus charge for origin traffic?
Yes. BytePlus documents $0.06/GB for data transferred from CDN to the origin server. In practice, that means cache hit ratio has a direct cost impact, not just a performance impact.
Are BytePlus CDN logs free?
Not always. Log Batch Delivery and Log Streaming are both paid, and each can add secondary downstream platform fees. BytePlus does document a free Log Download option, but it is slower and aimed at lower-urgency analysis.
Are 4xx responses billable on BytePlus CDN?
Yes. BytePlus explicitly says 4xx responses are billable and that the related requests also count toward billable request volume. This is one of the most important cost traps in the public documentation.
Are BytePlus savings plans always cheaper than pay-as-you-go?
No. Based on the published standard tables, the outside-Chinese-Mainland traffic savings plan is not automatically cheaper than first-tier pay-as-you-go in Europe or North America, and the request savings plan matches the published on-demand unit rate. Teams should verify the actual purchase-page pricing and not assume the savings-plan label guarantees lower cost.