For a continuous YouTube stream, compare regional outbound transfer and the complete monthly bill before comparing CPU and RAM. At a sustained 14 Mbps, a 30-day stream sends about 4.54 TB before protocol overhead, which is more than the listed 4 TB AWS Lightsail bundle allows.
That is a workload-specific comparison, not a universal plan recommendation. A VPS that relays an already encoded feed has different compute needs from one that encodes video, and the right choice also depends on region, stream quality and any extra traffic or redundancy you need.
Relay and encode are different workloads
A relay receives an already encoded stream and forwards it to YouTube. The video’s resolution, codec and bitrate are set upstream, so the VPS mainly has to maintain the connection and pass the data along. For this kind of workload, continuous outbound traffic and the route to YouTube matter before CPU or memory. You still need enough compute for the relay software and its operating system, but the research available here does not establish a suitable VM size for every relay application.
Encoding means turning source video into the compressed stream sent to YouTube. That work can be CPU-intensive, depending on the encoder, settings and source material. If you plan to encode on the VPS, do not choose a small plan solely because its traffic allowance looks generous. Test the actual encoder and content on a plan you can afford, then monitor resource use through demanding scenes and any transitions or overlays you intend to run.
This distinction also matters when you compare a cloud VPS with a workflow that sends a prepared file from elsewhere. YouTube’s recommended H.264 bitrate is 14 Mbps for 1080p30 and 8 Mbps for 720p30, but those figures describe encoding guidance, not a requirement that every channel stream at those rates. A devotional slideshow, an animated study room and a local news loop may have different visual needs and may be encoded differently. The live streaming guide to bitrate and resolution can help you decide what you are actually sending before you budget for its traffic.
Compare regional transfer allowances first
The advertised transfer allowance is not uniform across every location or plan. Hetzner’s traffic documentation lists 20 TB for EU CX, CPX and CAX cloud servers, while its listed US CX/CPX plans range from 1 TB to 5 TB, and Singapore plans from 0.5 TB to 5 TB depending on plan. These are listed allowances in the cited documentation, not guarantees that any particular region or product will remain available on the same terms. Check the exact plan and location you intend to order.
Lightsail uses monthly bundles with specified transfer allowances. AWS’s listed Linux bundles with public IPv4 include 1 TB at $5/month, 2 TB at $7/month, 3 TB at $12/month and 4 TB at $24/month. AWS counts both inbound and outbound transfer against the allowance, but says it bills only excess outbound transfer. Some regions have half the standard allowance, so a bundle that appears to cover a workload in one location may not do so in another. See AWS’s Lightsail bundles and pricing details and its data transfer allowance FAQ.
| Provider and location | Listed transfer context | What it means for a continuous feed |
|---|---|---|
| Hetzner EU | 20 TB for listed CX, CPX and CAX plans | Substantial room relative to a single 14 Mbps feed, subject to the selected plan and current tariff. |
| Hetzner US | 1–5 TB for listed CX/CPX plans | The lower allowances may be below a month of 14 Mbps traffic; check the specific plan. |
| Hetzner Singapore | 0.5–5 TB for listed CX/CPX plans | Plan choice is material, and the lower allowances can be below the example workload. |
| AWS Lightsail | 1–4 TB for the listed $5–$24 Linux bundles | The 4 TB example is below 14 Mbps for a full 30-day month before overhead. |
Treat this table as a comparison framework, not a live quote. The Hetzner figures are location- and product-specific; AWS’s allowance can vary by region. Both providers can change their terms, and the selected console or current official documentation should settle what applies when you create the instance. You can also compare the broader budgeting questions in this cloud service cost comparison for a 24/7 YouTube news channel, which considers a cloud setup alongside a desktop.
Compare the total monthly price
A low VM sticker price is not the same as a low recurring cost for a live channel. Start with the VM or bundle price, then include any public IP charged separately, transfer beyond the allowance, and any snapshots, backups or other services you actually use. Include tax where relevant to your bill, and use the same billing period and region when comparing options.
The public IPv4 difference is easy to miss. Hetzner’s Cloud overview says a cloud server does not automatically include a public IP and lists a Primary IPv4 at €0.50 per month excluding VAT. The same documentation says its Primary IPv6 IPs are not charged. AWS’s cited Lightsail bundle prices include public IPv4. That does not settle which one is cheaper for your case: a meaningful comparison must include the appropriate instance size, location, workload and expected traffic rather than pairing one provider’s entry price with another provider’s larger allowance.
The Lightsail prices above are AWS’s listed public-IPv4 Linux bundle prices in the documentation cited for this comparison, as listed on AWS’s site in September 2026. Hetzner’s 20 TB and regional ranges above come from its traffic documentation, which the research notes identify as last changed in 2024; they are not a monthly price. Check Hetzner’s current plan price and traffic terms before treating any figure as current. The €0.50 IPv4 charge is also a listed tariff, and should be checked against the current billing page for the selected account and product.
Do not add an overage charge unless your projected outbound use exceeds the applicable allowance, and do not assume all traffic is outbound. A relay may receive a source feed as well as send one to YouTube; Lightsail counts both directions toward its allowance, although AWS says excess billing applies only to outbound transfer. If you publish a backup feed or use the same instance for other work, include that traffic in the estimate too.
Estimate traffic for a continuous 14 Mbps stream
A steady bitrate can be converted into monthly data with a simple calculation: bitrate in megabits per second × seconds in the month ÷ eight bits per byte. For a 30-day month, the time component is 2,592,000 seconds. At 14 Mbps, that works out to about 4.54 TB using decimal terabytes. It is an estimate before protocol overhead, any second feed, retransmissions or unrelated traffic on the instance.
The same calculation gives about 3.24 TB at 10 Mbps and 2.59 TB at 8 Mbps. These are calculations for continuous transmission over 30 days, not provider-published usage statistics. Actual consumption can differ if the bitrate varies, the channel is not live for the whole period, or you run multiple streams. Build some room above the calculated figure rather than selecting an allowance that exactly matches it.
Fourteen Mbps is relevant here because YouTube’s encoder settings guidance recommends that H.264 bitrate for 1080p30. Its guidance gives 8 Mbps for 720p30. It does not mean every 1080p30 stream must use that rate, nor does it establish the right quality for every channel. The content, motion, codec and audience connection all influence the practical choice. Use YouTube’s current encoder settings and your own tests to set a quality target, then calculate traffic from the bitrate you actually configure.
If you are building a high-resolution music or image loop, do not assume that the resolution name alone tells you the traffic requirement. A 4K 60fps Tamil songs playlist workflow illustrates why the format and production choices need to be decided before the server budget. A higher stream rate produces more transfer whether the pixels are being encoded on the VPS or passed through by a relay.
Why the listed 4 TB Lightsail plan falls short here
The 4 TB Lightsail bundle is not enough for the specific example of a single uninterrupted 14 Mbps stream over a 30-day month. The estimate is approximately 4.54 TB before protocol overhead, so the shortfall is already present in the base calculation. Additional traffic or a second outgoing feed widens it. This is why the listed $24/month 4 TB bundle should not be described as covering that exact workload without qualification.
That conclusion has a narrow scope. It compares a 30-day month at a constant 14 Mbps against the listed 4 TB allowance. It is not a claim that all Lightsail streams exceed their allowance, that 4 TB is inadequate for every channel, or that the same allowance applies in every region. At 8 Mbps, the calculated base traffic is about 2.59 TB; at a lower average bitrate or fewer live hours, the traffic is lower. Conversely, an AWS region with half the standard allowance changes the comparison materially.
AWS says inbound and outbound data count towards the bundle, while only excess outbound transfer is billed. A relay receiving an upstream feed therefore needs to distinguish allowance accounting from excess billing, and should not simply treat all inbound data as irrelevant. Confirm the applicable region and billing explanation in AWS’s current documentation. If an adequate bundle requires moving to a larger tier, compare its full monthly cost with the alternative providers’ relevant region and plan rather than comparing only the $24 line item.
Choose a region, quality and plan
Choose the region based on the audience, your source feed and the actual route to YouTube ingest, then verify the available allowance in that region. If most of your viewers are in India, proximity to viewers can matter for playback experience, but the VPS’s outgoing connection is to YouTube’s ingest endpoint; this comparison does not establish which region will give your channel the best route. Test from the location you intend to use and avoid assuming that a region’s transfer allowance predicts stream stability.
Next, settle the bitrate and number of feeds. For each continuous feed, calculate monthly transfer from its configured bitrate and operating hours, then add headroom for protocol overhead and other traffic. YouTube recommends leaving 20% of bandwidth headroom in its streaming tips; that is guidance for connection capacity, not an instruction to increase the monthly data estimate by a fixed percentage. Keep these two questions distinct: capacity headroom helps a connection absorb variation, while transfer allowance is a billing quota.
For a relay, prioritise an allowance that covers the feed and a network path that works in your tests. For encoding, benchmark the chosen encoder with your real source and settings before committing to a longer term or moving a channel that cannot tolerate an interruption. There is no evidence here to name the cheapest adequate CPU/RAM plan for an unspecified encoder. A small business loop that has already been encoded might be an easy relay; a channel compositing scenes and encoding live on the VPS needs its own compute test.
Your operating preference matters too. Lightsail presents a simple bundle ladder with IPv4 included, which can make billing easier to read. Hetzner’s listed EU transfer can be generous relative to the example traffic, but the allowance differs by location and a public IPv4 is separately charged in its documentation. The better fit is the plan whose regional allowance, required compute and complete bill suit your own workload—not a provider label in isolation.
Verify the operational details before committing
Check the exact product, region, transfer allowance and price on each provider’s current page or console. Confirm whether the chosen plan includes IPv4, what outbound overage costs, whether snapshots or backup retention add charges, and whether taxes are included in the displayed price. For AWS, verify whether the named region receives the standard or reduced allowance. For Hetzner, check the current traffic table for the specific plan family and location; do not assume the EU row applies elsewhere.
Then test the path and the stream itself. YouTube’s guidance recommends constant bitrate, a two-second keyframe interval (not over four seconds), and RTMPS. It also advises reliable connectivity and monitoring stream health. Those recommendations do not guarantee that a particular VPS will stay connected, so run a test stream from the intended region, inspect YouTube’s health information and verify that your process reconnects if the connection drops.
Availability commitments are narrower than an end-to-end live broadcast. Hetzner’s Cloud and vServer service agreement describes a commercially reasonable 99.9% monthly availability target for each Cloud Server under its contract definitions and exclusions. That is a contractual service term, not a promise that your encoder, network path, stream key or YouTube ingest will remain healthy. Plan for alerts, restart behaviour, reconnect testing and a way to intervene if a stream stops. YouTube explicitly cautions that a connectivity disruption can break a stream.
A cloud VPS can also leave you responsible for setup and maintenance. If you need to manage the encoder or relay process yourself, document its start-up, log checks, key handling and recovery steps before handing the channel over to an unattended schedule. If your main problem is that a home computer must stay switched on, a hosted file-to-stream workflow may remove that particular burden: StreamNeo takes an uploaded video and runs it as a YouTube live stream, so you do not need to keep your own computer running for that broadcast.
Before committing, compare the operating options on the pricing page. When the file and channel are ready, start free — 24-hour trial, no card.
FAQ
Is 1 TB enough for a 24/7 YouTube stream?
It depends on the bitrate and how much of the month you stream. At 8 Mbps, a continuous 30-day feed calculates to about 2.59 TB before protocol overhead, so 1 TB would not cover that example. Calculate from your actual configured bitrate and check the selected region’s allowance.
Is the 14 Mbps figure a requirement for every 1080p stream?
No. YouTube lists 14 Mbps as its recommended H.264 bitrate for 1080p30, and 8 Mbps for 720p30. These are recommendations for those settings, not a universal requirement for every channel, codec or visual style.
Does a Hetzner availability target mean my YouTube stream will stay live?
No. The contract’s availability language applies to the Cloud Server within its defined scope and exclusions. Encoder failures, reconnect behaviour, the route to YouTube and YouTube ingest are separate parts of the broadcast, so test monitoring and recovery rather than treating the VM term as end-to-end uptime.
Should I encode video on the VPS or relay an existing stream?
A relay forwards an already encoded feed and is mainly a transfer and connectivity workload; encoding adds processing demands that depend on your software and settings. If you encode on the VPS, test the real source and encoder on the intended plan. If you only need to keep a prepared video looping, compare that workflow with the operational effort of managing a VPS.