Ultra-low latency
Ultra-low latency is a class of live video delivery where the delay from camera to viewer is under about one second. Reaching this level usually requires real-time protocols such as WebRTC rather than standard segment-based HLS or DASH. The result feels close to a video call, where people can talk and react with almost no lag.
Why it matters
Some products only work if every viewer sees the same moment at the same time. Betting, auctions and live Q&A break when one viewer is five seconds behind another. Ultra-low latency makes these real-time products possible, but it costs more to run at scale than standard streaming.
How it works
- The sender captures and encodes each frame and sends it immediately, without grouping frames into segments.
- A real-time protocol such as WebRTC carries the media over fast, connection-based transport.
- Media servers forward the stream to each viewer instead of caching files on a standard CDN.
- The player keeps only a tiny buffer, so frames appear almost as soon as they arrive.
What is the difference between low latency and ultra-low latency?
Low latency usually means about 2 to 5 seconds of delay, while ultra-low latency means under about one second. Low latency is reached by trimming segment-based HLS or DASH. Ultra-low latency usually needs a real-time protocol such as WebRTC. Note that some platforms use the label loosely: YouTube's setting called ultra-low latency targets under 5 seconds.
Do I need ultra-low latency for my live stream?
You need ultra-low latency only when viewers act on what they see in real time, such as betting, bidding or talking with the host. For most sports, concerts, classes and webinars, 2 to 5 seconds is enough. Choosing ultra-low latency when you do not need it raises cost and limits audience size.
Can HLS reach under one second of latency?
No, HLS generally cannot reach under one second of latency, even in its low-latency form. LL-HLS usually lands around 2 to 5 seconds because players still need small parts and a short buffer. To go under one second, teams typically use WebRTC, often with WHIP and WHEP for ingest and playback.
Where it's used
Betting and live auction apps use it so odds and bids match the picture for everyone. Live event platforms bring remote guests on stage, and interactive shopping hosts answer viewer questions without awkward pauses.
By the numbers
- WebRTC became an official W3C Recommendation and IETF standard on 26 January 2021. (WebRTC becomes a W3C Recommendation and IETF standard (W3C))
- The IETF published WHIP, an HTTP-based protocol for sending WebRTC media to streaming services and CDNs, as RFC 9725 in March 2025. (RFC 9725: WebRTC-HTTP Ingestion Protocol (WHIP))
- YouTube's setting named Ultra-low latency gives most viewers under 5 seconds of delay and does not support 4K resolution. (Understand live streaming latency (YouTube Help))
