GPU AI/ML today! NVIDIA A100s & H100s for €2.06/hRead more
Under attack?





Why Gcore


Low Latency

Live Streaming

Infrastructure for live streaming.

Scale to 100+ million viewers and beyond

  • €0

    Adaptive bitrate encoding at no cost

  • 0.3 to 4 sec latency

    Scalable and reliable streaming


    Simple dashboard, SDK and API

Workflow ready

Scale with us if you have your own video server and apps. And use our infrastructure with open source demos if you need to build a new video service from scratch. RTMP, SRT, WebRTC or multicast broadcasting of your streams to million of viewers.

Workflow ready
Scroll horizontally to see the diagram

Simple steps to start

Step 1
Create a stream

Get a stream key for PUSH or PULL in the Dashboard

Create a stream

or use API to get a secret

Request: { "stream": { "auto_record": true,"active": true} } Response: { "id": 309827,"push_url": "srt://","backup_push_url": "rtmp://" }
Step 2
Send a stream

Gcore receives RTMP, SRT, HLS, WebRTC, UDP Multicast, and other protocols, which are supported by most broadcast software/hardware as well as open source software for mobile applications.

    Web studio in just browser
    Streaming from mobile apps (open source demo & howto)
    OBS and vMix parameters
Step 3
Watch a stream

We prepare adaptive bitrate for devices and bandwidth automatically. Use our HTML-player for free or insert manifest into your player to play LL HLS, MPEG-DASH CMAF streams.
src="" allow="autoplay; encrypted-media" allowfullscreen />

Ready to get started?

14 days free trial

Fast delivery

You get cutting-edge streaming technology without investing in expensive infrastructure. Delay in delivery to the end viewer no more than 4–5 seconds. At the same time, high quality audio and video up to 4K/8K is maintained.


For video content that is very sensitive to delays (sports, gaming, news, auctions, interactive shows), we use top-notch technology to minimize such disruptions: LL HLS, MPEG-DASH CMAF, HESP and low-latency CDN


Video Capturing
    SRT, RTMP input
    MPEG2 UDP multicast input
    WebRTC input from browser
    PUSH & PULL input
    Backup input of live streams
    Adaptive mobile RTMP
    Video conferencing input
    Realtime encoding
    Adaptive bitrate
    Video hosting
    Low Latency transcoding
    Video 360
    Server-Side Video Insertion (SSAI)
    Statistics and analytics
    Pure CDN delivery
    Low Latency Streaming
    Restreaming to social networks
    Video protection
    Playback on iOS, Android, Windows, Mac, Linux, Set Top Box, Smart TV, and game consoles
    HLS, MPEG-DASH, fragmented MP4
    HTML5 player

Frequently Asked Questions

You need a streaming software/videocam that sends streams in rtmp/srt.
Or if you’re a pro, you may have a pre-set up server from which we can extract your srt/rtmp stream.

After we get your stream, we transcode it in the qualities you need and to the http-compared protocol (HLS/DASH) which are suitable for web-players and CDN.
Then we deliver this stream to your audience anywhere in the world using our CDN.
We provide you with an iframe of our player to embed into your application. But that’s not mandatory. You can use any other hls.js/dash.js compatible player to place our playlist in.

Here is a step-by-step guide explaining what settings you need to set up in the Control Panel.

We accept any kind of software or videocams. It doesn’t matter what you use—free OBS or Entreprise LiveU, or just a videocam with a pre-setup software.

You just need to make sure that your software works with rtmp or srt. Most modern softwares surely accept these protocols.

If you have something very specific, just contact us.

Sure, you can. This protocol is much more stable than RTMP—it doesn’t drop a connection if there’s something wrong with provider routes.
We recommend all our customers use SRT. But RTMP also works as the most common streaming protocol.
Check, please, whether your software supports SRT, then send a stream to us in that protocol.

From 1 to 100,000,000 or even more. There’s no end-point.
We use our own Edge Network with PoPs all over the world united into redundant clusters, so our edge-servers share load between each other. This means that viewers are separated between lots of servers based on the following:

  • Geography/Topography – which means users get the stream from the nearest location (both geographical and providers routes)
  • Overload – if one server in a location has a full channel, the next viewer from the same geographical point will be balanced to the neighboring server.

And of course, if one server is out of work for any reason, traffic will be re-routed to another server in the same cluster. This means viewers will never get drops in the Live Stream.

No, you don’t.
We have our own CDN with servers all over the world. You don’t need to look for another provider for your stream’s delivery, as it’s all built-in for you. Our CDN is preset for streaming by default if you’re using our Streaming Platform.
All the settings for caching playlists and chunks are made by us, based on the experience and average stream parameters. We also cache streams not in the HDD/SDD of our servers but in the Operative Memory, which means that content will be delivered to viewers faster and with no freezes.

The information will be here soon.

We use our own infrastructure. This means we can set-up the ingester/transcoder for you in any location where Gcore has Cloud service.
By default, we have ingesters/transcoders in Ashburn (USA), Luxembourg (to cover Europe) and Singapore (to cover Asia). Of course, they’re united into clusters for redundancy and for overcoming huge loads.
Those locations are usually enough for the streamers from the USA/Europe/Asia. And we deliver these streams using our CDN. So, if your streamer is located in the USA and the viewers are in Europe, viewers will get the stream from the European servers.
But If you still want something closer to your streamers, check the map and send a request to us.

We have the following latency for the following protocols:

  • Reduced latency HLS (the most common for any cases) is 6-8 sec
  • Low latency DASH (the most stable LL, but doesn’t work in native iOS AVPlayer, although it works in Safari, iOS, with our player) is 3-4 sec
  • Apple low latency HLS (LL working on any device including iOS) is 4 sec
  • HESP (special HTTP-based protocol, works only in the specific player) is less than 1 sec

We also have WebRTC for use with VideoCalls – real-time latency. But that technology isn’t compatible with CDN. So, it’s better to use WebRTC for real-time communication, not for high-quality streams. Read more here.

We don’t charge for our transcoding in the basic protocols (that means quality up to 1080, HLS/DASH/LL-HLS).
You pay only for the minutes your viewers watch the stream.
Price for 1 min of watching is €0.001. For example, if you had a 1-hour stream and 10 viewers watched it from the beginning till the end—that would be 60mins * 10viewers * €0.001 = €0.6.
We charge per month. That means if during one month there was only one 1-hour stream with 10 viewers—you pay only for that. If in the next month you expand to ten 100-mins streams with 100 viewers watching each stream from the beginning till the end, you’ll pay 100users * 100mins * 10streams * €0.001 = €100.
It depends on your consumption, and you don’t pay more than your viewers have watched.

Contact us to get a personalized offer