This guide covers performance recommendations for creating real-time video pipelines.
Pipeline pricing follows a transparent, usage-based model that reflects the Livepeer network’s open marketplace. This guide explains how pricing works and what to expect when using pipelines in your applications.
You pay only for the time your streams use a pipeline. Pricing is calculated per minute of video processed, with costs starting when a stream connects and ending when it disconnects. It is scaled by pixels, so if you use a high-res video you will incur greater costs.
Copy
Total Cost = Minutes Streamed × Pixel Multiplier x Pipeline Rate
Pipelines run on a network of infrastructure providers (called Orchestrators). Each provider can set their own rates for running specific pipelines. When you start a stream, we automatically select the most cost-effective provider that meets your pipeline’s performance requirements.You can also choose a specific provider if you prefer. This might be useful if you:
Need guaranteed infrastructure in specific regions
Right now, you will need to manage your own costs. In Q1, we will begin rolling out a suite of features to help you understand and control your spend.These will include:
Q: What happens if I run out of funds?
A: Your streams will continue if you have automatic refills enabled. Otherwise, they’ll be paused until you add more funds.Q: Can pricing change during a stream?
A: Yes. However, if you choose the “Best match” provider selection option, it is guaranteed that your rate will not increase.If you choose to run a pipeline with a specific Livepeer orchestrator, the rate you see when starting a stream is guaranteed for its duration unless the orchestrator fails. In that case, the stream will fall back to the “Best match” option as described above.