Agora is a useful CPaaS company for voice and video chat, messaging, and interactive live streaming. They can scale from one to many with sub-second latency, and they have a customizable user interface with plenty of APIs to keep developers busy embedding audio and video call features into communications systems. They support industries from social media, online dating, mobile gaming, e-commerce, and telehealth. Agora’s WebRTC-enabled Real-Time Engagement PaaS runs on their proprietary Internet network set up in 200 worldwide datacenters. Agora offers well-built, fully functional multiple SDKs and they are good at point-to-point communication.
Like any service, Agora has some drawbacks, particularly with streaming. Although video to up to a million passive viewers can be streamed over the Agora network (although we’ve yet to see examples of this scale), the platform can only support interactive streaming for a maximum of 17 people. Their recommended maximum video resolution with interactive engagement is 240p on a one-to-many call, and 360p on a one-to-one call. The maximum resolution they recommended on passively streamed video is 720p. While you can set higher resolutions, the costs associated can be detrimental to your monthly Agora charges. We’ve heard from customers who have switched (or are in the process of doing so) from Agora to Red5 Pro that the quality of HD video on Agora overall is not great even though the company claims they support it.
There are several factors that go into selecting a live streaming provider, namely latency, functionality, price, quality, and scalability. When considering alternatives, there are a variety of similar options that may better suit your needs. We here at Red5 Pro know of a great choice, but we encourage you to look at all your options.
Here are some of the leading Agora competitors that provide live video streaming services, in no particular order:
ANT Media’s inexpensive price point may tempt those looking to save money. Their non-hosted solution is $49 per server instance, and their hosted offering appeals to teams lacking the resources to maintain their own servers. According to ANT’s website, because of WebRTC, they can provide an expected latency of 500 ms or less. On the surface, this all sounds good, but do not cut corners.
Scalability is ANT Media’s kryptonite. Their Large Enterprise Instance hosted package limits a user to 8 concurrent broadcasters and 400 high latency subscribers (HLS). WebRTC subscribers are confined to only 300 clients. This low cutoff number deeply restricts growing an application. But...ANT Media does have a Custom Scalable Cloud with a flexible number of publishers and viewers.
Interesting fact: ANT Media does not always stay up to date. ANT Media built their platform on top of our Red5 open-source software. However, we at Red5 Pro retain control of the open-source Red5 code. So we update the open source as new optimizations and features are added. This minimizes conflicts and bugs when the Red5 open source and Pro systems maintain parity and sync together seamlessly.
ANT Media has difficulty ensuring consistent functionality likely because they do not contribute to Red5 open source. Plus, they are limited in innovating new features—they (so far at least) only copy the features we come up with. Without progress, this lack of originality results in slow updates and product features, or even dysfunction.
Finally we’ve heard over and over again from customers who’ve switched from ANT Media to Red5 Pro that their support is lackluster.
Another Agora alternative, Millicast, is a WebRTC-based hosted solution. They offer a combination of pre-packaged hosting convenience, with WebRTC’s low latency of 200-500 milliseconds, making it real-time streaming.
This increased convenience is at a price. As users completely rely upon third-party architecture, one is forced to accept any changes the third party makes. This is called a service trap—a user is bound to the service their application was built on. Like Agora and other purely hosted solutions, customization is stunted with their closed back-end infrastructure. One risks being blocked from implementing one's own server-side logic, like transcoding, integrating with AI, implementing custom authentication, and pushing a stream out to other processes. Rather than creating an application specifically tailored to one’s needs, a user is confined to a pre-built sandbox. Basically if Millicast, Agora, TokBox, etc. don’t support what you need out of the box, you are out of luck. One approach to avoid this is to use plug-in based architecture hosting-agnostic solutions like Red5 Pro.
As is also the case with Agora, another limitation of Millicast’s hosted model means that your app’s infrastructure is shared with other companies’ apps. At first glance this might not seem like a major problem, however, we have heard reports of Millicast servers being blocked by network-based child content protection filters. This is due to many Millicast customers being in the gambling space and/or other inappropriate non-kid-friendly businesses. If your app is geared towards education, for example, this may be a red light for you.
It’s worth noting that Millicast is expensive. Their hosting packages include a limited amount of data, and they charge if a user goes over. For example, their “Pro” plan is listed as $2,495 per month with 40,000GB of data included. That’s a decent amount of data and considering that it’s hosted and mobile SDKs are included so, ok, that’s fair.
Compared to Red5 Pro, the Growth Pro plan and mobile SDKs with DigitalOcean hosting cost $1,238 a month. That costs half of what Millicast charges, and find a further discounted rate with a Red5 Pro annual plan.
With a long history in the video streaming sector, Wowza has powered a lot of customer streaming solutions. This large following has made Wowza a trusted media provider. Their growth has been driven by the fact that they support a variety of ingest types. Accordingly, a vast percentage of their customers use the Wowza Streaming Engine for a single purpose: serving as an ingest point (an origin server) to deliver to a CDN. After taking in a number of streaming protocols, the origin then converts (often transcoding) the streams for HTTP based CDN delivery. Repackaging streams for CDN delivery is actually a good use of Wowza’s software assuming you don’t care about latency.
About a year ago, Wowza declared end-of-life status for their mobile SDKs, thus limiting their mobile capabilities. If you want to create a native app to run on Wowza, you will have to build it from scratch without the convenience of an SDK built for their platform. Note that unlike Agora or Red5 Pro, Wowza does not have an SDK for browser-based applications either. This means users will spend lots of time working with the complex WebRTC API directly doing things like manually “munging” SDPs to connect to a Wowza system.
The large size of Wowza (recently taken over by a private equity firm) reflects their ability as a company to distribute a workable product. However, being large can make a company unwieldy. Important shifts in the tech-sector can happen quickly. Anticipating and reacting to those shifts can have a meaningful impact on the future effectiveness of your product.
One such shift came with the introduction of the Web Standard WebRTC for live streaming use cases. Designed for low latency, WebRTC is meant for browser-based live streaming due to a highly functional protocol. Wowza, like the original Red5 was built as an alternative to the Flash Media Server delivering video over RTMP. With Flash reaching end of life support, WebRTC became the obvious replacement for real time latency video streaming. In the face of this ever-increasing urgency, Wowza’s efforts to fully implement WebRTC seem to be falling short.
For any kind of scale, Wowza relies on CDN-based video delivery, thus they struggle with high latency of around 2 seconds via Apple’s LL-HLS (at the very lowest). This is nowhere near the sub-500 ms delivery that WebRTC has produced on other streaming solutions. 2 seconds is too high for live, real-time interactive experiences.
Also, using Wowza Cloud may entail being locked-in to their network. Without a clear way of porting an existing application to a different hosting provider, a user is forced to adopt Wowza network changes.
Lastly, Wowza has a customization limit. They created a general-purpose platform with basic functionality, making their platform highly accessible. However, that simple functionality means custom feature development or even modifications may be difficult to implement or receive necessary support.
Another convenient hosted solution and Agora alternative is DACast. DACast has an HTML5 player that supports mobile devices and browsers. VOD is also supported, which lets clients insert ads. However, DACast lacks a native SDK for mobile-specific development with a native application. This can have a negative impact on feature performance and the user interface.
Although DACast appears to have mid-level pricing, their pricing plans are confusing, and are tough to compare against other platforms. Also, prepare to do some interesting math, as DACast only lists the monthly plan breakdown.
When it comes to low latency, DACast is hampered by its build. They use CMAF, which requires segmenting the stream into small chunks, producing unfortunate high latency. In fact, their website states a latency of 10 seconds (although realistically CMAF should get down to about 3 seconds). As mentioned, a platform that uses seconds to measure latency cannot produce low-latency video streaming.
Another major Agora competitor, TokBox is useful for live streaming video. Their relatively easy-to-use interface allows a user to quickly set up for basic plug-and-play use. Their SDKs are highly effective at point-to-point communication. Their basic functionality and fast setup time makes TokBox ideal for establishing POCs during large or small events.
Despite being acquired by Vonage, most still refer to it as TokBox regardless of its official new name, “Vonage API.”
TokBox can suffer from some pitfalls. Specifically in regards to extra features, customization potential, and scaling costs. They charge per stream per minute. This means that as you gain more users your growth can actually hurt you: the more stream will exponentially increase costs. Don’t let success harm your business because of this.
For example, if each day a user has 3 broadcasts lasting 30 minutes with 100 subscribers each:
100 subscribers x 30 minutes x 3 broadcasts is 9,000 streaming minutes a day.
5 days a week for 4 weeks is a total of 180,000 streaming minutes a month.
The first 2,000 minutes is $9.99 meaning 178,000 additional minutes will cost 0.0045 per minute. The grand total = $810.99 per month.
Conversely, Red5 Pro with DigitalOcean hosting costs $427.05 per month. Even without the annual plan savings, that’s almost half the price of TokBox.
Additionally, TokBox charges more for their “advanced” features. Their requirements for what qualifies as advanced could and should be looked at as basic features. For example, recording and interactive broadcasts (which do not come with a guaranteed low latency), are their advanced features. Plus, using more than 2,000 connections will force a change to CDN delivery. As we’ve covered before, CDNs cause higher latency and other issues.
Similarly, after 3,000 viewers, the platform switches to HLS. The concern with HLS is that it will add many seconds of latency, proving TokBox cannot deliver real-time streaming at scale.
Agora competitor Mux, has created an easily managed service designed for basic streaming applications. Created with developers in mind, their solid documentation gives those with a basic understanding of software dev skills a good outlet to set up a simple streaming video platform regardless of specific video experience.
Like Agora, Mux also bases their pricing on usage. In addition to the network bandwidth accrued through video delivery, they also charge for live video encoding as well. At just $0.07 per minute of video encoding and $0.0013 per minute of video streaming, they are a less expensive alternative to Agora but still quite pricey.
Where Mux really shines is with their innovative multi-CDN approach. This optimizes stream distribution across different regions and avoids being locked in to a specific CDN. That said, users cannot stream outside of their network, and as we get into next, you can forget about real-time latency.
As innovative as the multi-CDN approach is, it will still face the downside of CDNs, with all the video delivery performed through high latency HTTP based protocols. In Mux’s case, they use HLS, which results in the highest level of latency: anywhere between 10-30 seconds. Even if they are using Low Latency HLS (LL-HLS), which they haven’t offered yet, that would still produce a latency of 2-5 seconds. With a delay that high, any sort of live, interactive experience will not work.
Limelight is a traditional CDN focused on live streaming that’s lately attempting to get into the real-time game. Their Real-time Live Streaming (RTS) platform is a managed solution, which makes it straightforward to configure everything. With their mobile SDKs, users can build native apps. By handling all the server infrastructure, the setup process is relatively easy. With consistent performance, and reliable content delivery, Limelight works well.
The RTS feature is an offering that lets Limelight support true low-latency streaming, so subscribers and broadcasters respond to events in real time. Limelight used Red5 Pro’s software to build their RTS platform. Through Red5 Pro, they achieved sub-500 milliseconds of latency. Despite this real-time latency, there are still limitations. Limelight did not enable current Red5 Pro features, which blocks full functionality.
Limelight is not forthcoming with their pricing structure, though rumor has it they are expensive. Their business is experiencing issues at the moment and going through some changes that may affect their customers. It’s also not clear whether Limelight plans to continue to support RTS moving forward.
It’s said that Agora alternative Phenix is a well-functioning, reliable streaming service. One of their most notable features is a fast time to first frame. However confirming this has been tough, as they do not offer any live demos, free trials, or exposed example code. Although they’ve partnered with Verizon, there is nothing on Phenix’s site showcasing their pricing. If you want more information, you will need to contact them.
We couldn’t discuss live streaming solutions without coming up with at least one open source alternative to Agora. Kurento is a solid choice for live streaming because their software integrates WebRTC. They are completely free open-source (it is the only open-source solution on this list).
Nonetheless, just as convenience has a price, so too does a completely free, DIY pre-solution. It may not cost money, but a user makes up for that with the cost of time and effort to build a full-featured app from scratch. If you know what you’re doing, Kurento can provide a good backbone.
There are many other open-source WebRTC options available such as Pion, Jitsi, Media Soup, and Janus. While these others are good, Kurento is one of the most widely used and best known of the many WebRTC-based media servers. That said things are changing quickly, and Pion is seemingly gaining momentum.
As our tagline says, we provide live streams to millions of concurrent users with milliseconds of latency. Like others on this list, we use WebRTC to ensure video delivery with sub-500 milliseconds of end-to-end latency. However, one major differentiator is that we can further extend that real-time latency to millions of concurrent connections. We also support many other real-time protocols like SRT, RTSP, and MPEG-TS making it completely interoperable with WebRTC. Our hosting-agnostic solution leverages the cloud provider of your choice, addressing lock-in concerns and avoiding a service trap. This gives users a great degree of flexibility over back-end architecture. Plus, the rest of the software is fully customizable too. With full-featured SDKs for iOS, Android, and Linux, you can get the same performance across mobile devices.
As a highly customizable solution, for perfect functionality, it’s necessary to perform some additional configurations. But no worries, we’ll never let you struggle on your own. Our detailed documentation will get you started, and our responsive technical support team ensures that everything works as expected.
Though we don’t currently provide the convenient option of a hosted solution, we do have one in beta, and it’s set to launch this fall. Those looking for a hosted solution right now, we offer an Enterprise Plan in which we set up and manage an account for you. Enjoy security and flexibility of running your own servers minus additional maintenance.Explore our ultra low-latency video streaming demo and sign-up for a 30-day free trial. Contact us with any comments: send an email to firstname.lastname@example.org or schedule a call.