H.264 Was Finalized in 2003. It Still Carries More Than 80 Percent of Internet Video Today.
In May 2003, a committee called the Joint Video Team published a specification that most people had never heard of and would never read. The document described a video coding standard formally designated H.264, also known as Advanced Video Coding or AVC. The Joint Video Team was itself a partnership between two standards bodies: ITU-T's Video Coding Experts Group and ISO/IEC's Moving Picture Experts Group. The two groups had formed in December 2001 with a focused mandate: build a codec that could deliver broadcast-quality video at half the bitrate of the then-current standard, MPEG-2. They achieved exactly that. What nobody fully predicted was that this standard, finalized in a Geneva committee meeting, would still be moving the majority of the world's internet video more than two decades later.
The timing of H.264's release was not accidental. The internet in 2003 was not yet a video medium in any meaningful sense. Dial-up was giving way to broadband, but bandwidth remained precious, and video files were absurdly large by the standards of what could be delivered reliably. H.264's core achievement was compression efficiency: it could encode the same visual quality as MPEG-2 at roughly half the data size. That ratio was enough to tip the economics of online video into viability. YouTube launched in 2005 and had embraced H.264 encoding by 2009. Apple shipped the original iPhone in 2007 with H.264 hardware decoding built in, making it the default assumption for mobile video for years to come. By the time streaming services became household fixtures, H.264 was so deeply embedded in device silicon, encoding pipelines, and browser implementations that it had become essentially invisible, the thing video just was.
According to Bitmovin's Video Developer Report published in December 2024, H.264 remains the format used by 79 percent of video industry developers, and achieves 98.23 percent browser compatibility across desktop and mobile platforms. No other codec comes close to those numbers. Understanding how H.264 got there, why its successors have struggled to displace it, and what is finally beginning to challenge it reveals a great deal about how technical standards actually spread through the world, and why the video files most people produce today are still larger than they need to be.
The Architecture of Dominance
H.264's technical design was sophisticated for its time, introducing a range of encoding tools including variable block-size motion compensation, in-loop deblocking filters, and multiple reference frame support. These collectively produced much better compression than previous codecs. But plenty of technically superior codecs have failed to gain traction. What made H.264 different was a coordinated approach to licensing.
Patent rights to H.264 were pooled under an organization called MPEG LA, which became Via Licensing Alliance in April 2023. Rather than allowing individual patent holders to demand separate royalties, MPEG LA offered a single license covering the entire patent pool. In August 2010, the organization announced that royalties would not be charged for H.264 encoded internet video free to end users. That decision effectively made H.264 free for the vast majority of online video distribution, removing the last significant barrier to adoption by streaming platforms, browser vendors, and device manufacturers. Within a few years, H.264 decoding was baked into the hardware of essentially every smartphone, smart television, laptop, and set-top box sold anywhere in the world. That hardware support created a self-reinforcing cycle: because every device could decode H.264, every encoder targeted H.264, which meant every device needed to keep supporting it.
The codec also benefited from an unusual degree of stability. Unlike software products that receive constant updates, H.264 was a fixed specification. Implementations improved over time, but the bitstream format was set in 2003. A video encoded in H.264 in 2009 plays on a device manufactured in 2026 without any compatibility work. That kind of guaranteed backward compatibility is extraordinarily rare in technology and has genuine economic value for anyone managing a video archive at scale.
When a Better Codec Ran Into a Licensing Wall
H.265, formally known as HEVC (High Efficiency Video Coding), was finalized in 2013 by the same joint team that had created H.264. It delivered on its technical promises: at equivalent visual quality, H.265 files are roughly half the size of their H.264 counterparts. For a content provider streaming millions of hours of video daily, that compression gain translates directly into measurable bandwidth costs and storage expenditure. The problem was not technical. It was legal.
Unlike H.264's relatively unified licensing structure, H.265 attracted multiple competing patent pools, each claiming royalties from different sets of patents. By the mid-2010s, at least three separate licensing organizations were asserting rights over H.265: MPEG LA, Via LA, and HEVC Advance, with a fourth entity, Sisvel, joining later. Content providers and device manufacturers faced the prospect of negotiating and paying multiple overlapping licenses for the same codec. Many simply chose not to. H.265 hardware decoding eventually appeared in smartphones and smart televisions, but adoption among streaming platforms lagged well behind what the technical merits alone would have predicted.
Google took a different path. Rather than working within the patent-pool system, Google developed VP9 as an open, royalty-free alternative and began deploying it on YouTube in 2013. VP9 achieved compression efficiency comparable to H.265 without the licensing complexity. It also established the philosophical template for what followed: a coordinated industry effort to build something better outside the patent system entirely.
The Open Alternative That Is Finally Making Ground
The Alliance for Open Media formed in 2015, bringing together Google, Mozilla, Microsoft, Apple, Amazon, Netflix, and Intel with a shared goal of developing a royalty-free video codec that could outcompete H.265. The result was AV1, released in 2018. AV1 delivers genuine improvements over both H.264 and H.265. At equivalent visual quality, AV1 files run approximately 20 to 30 percent smaller than H.265, and in content categories like animation and screen recordings, the gap reaches 40 to 60 percent. A ten-minute 4K video that compresses to 1.5 gigabytes in H.265 might compress to roughly 900 megabytes in AV1. These are not marginal differences.
Hardware decoding support for AV1 has expanded significantly in devices released after 2021. Major streaming platforms including Netflix and YouTube began deploying AV1 for supported devices, and modern smartphone processors include dedicated AV1 decode hardware. The royalty-free status removes the specific obstacle that stalled H.265 adoption, meaning platforms can deploy AV1 without navigating competing patent claims or accumulating licensing costs at scale.
Yet the transition remains incomplete. AV1 encoding is computationally demanding. Encoding the same video in AV1 takes substantially longer than encoding it in H.264 or even H.265, which matters for anyone processing video on consumer hardware or publishing time-sensitive content. A meaningful share of devices in active use still lacks hardware AV1 decoding acceleration, meaning playback can be demanding on older hardware. The result is that AV1 is genuinely taking hold at the infrastructure level of major platforms, while consumer-facing defaults remain rooted in H.264.
Why Most People's Video Files Are Larger Than They Need to Be
The practical consequence of H.264's dominance is specific and common. Most video content that ordinary users produce, from family recordings to screen captures to short clips, arrives as H.264 files. Cameras, phones, and screen-recording software default to H.264 because it plays everywhere. But H.264 at typical camera settings is often far from optimal. A camera recording at a high bitrate for archival quality produces files that are far larger than necessary for sharing or streaming at the same perceptible quality level.
Many of these files contain more data than any viewer's screen or connection can actually benefit from. Converting or recompressing a video while targeting a more appropriate bitrate for its intended use can reduce file sizes by 50 to 70 percent with no visible quality change under typical viewing conditions. The codec has not changed, but the use case has moved far beyond what the camera manufacturer's defaults were designed for. A 4K recording at 100 megabits per second that will be watched on a phone screen over Wi-Fi is carrying roughly ten times more data than the viewing context can use.
Conclusion
This gap between what most video files are and what they could be is the practical residue of H.264's dominance: the format is everywhere, the defaults are conservative, and most data in most video files is not serving the viewer. ToolHQ's Video Compressor at https://toolhq.app/tools/video-compressor handles this at the compression stage, processing files securely on the server and deleting them immediately after conversion, so no local software installation is required.
For related tasks including audio compression and animated GIF optimization, ToolHQ also offers dedicated tools for MP3 and GIF files. The story of H.264 is ultimately the story of how technical standards become infrastructure, persisting long after better options exist because they are embedded in every camera, browser, and streaming pipeline on the planet.
Frequently Asked Questions
Why did H.265 fail to replace H.264 despite better compression?
H.265 attracted multiple competing patent pools simultaneously, creating a licensing nightmare. Content providers faced royalty claims from MPEG LA, HEVC Advance, and Sisvel at once, making adoption too legally complex compared to H.264's single unified license.
How much smaller are AV1 files compared to H.264 at the same quality?
AV1 typically produces files 40 to 50 percent smaller than H.264 at equivalent visual quality. For animation and screen recordings the gap can reach 60 percent. A 1.5 gigabyte H.264 video might compress to under 800 megabytes in AV1.