AV1 in 2026: Adoption, Hardware Support, and What It Means for Your Pipeline

Three years ago, AV1 was a promising but impractical codec — excellent compression, terrible encoding speed, limited hardware decode support. That has changed. AV1 is now the primary codec for the largest video platforms on Earth, hardware decode support covers the majority of active devices, and encoding tools have matured to the point where AV1 is a realistic option for any video pipeline.

Here is where things stand.

Last updated: March 2026

#AV1 by the numbers

The adoption numbers tell a clear story:

• Netflix: AV1 serves approximately 30% of all streams, prioritised on devices with hardware decode support. Netflix has been the most aggressive major platform in pushing AV1 adoption, driven by the direct CDN cost savings from smaller files.
• YouTube: Over 75% of video playback on YouTube now uses AV1, up from roughly 50% in 2024. Google has been deploying AV1 aggressively across Chrome, Android, and smart TV apps.
• Meta: AV1 powers more than 70% of video playback across Facebook, Instagram, and Messenger. Meta’s adoption accelerated after they completed their AV1 encoder optimisation work in late 2024.
• Device certification: Among large-screen devices (TVs, set-top boxes, gaming consoles) certified between 2021 and 2025, approximately 88% include AV1 hardware decode support.

These are not projections or roadmap items. This is current production traffic. The largest video distributors in the world have already made AV1 their default codec for supported devices.

#The compression advantage

AV1’s value proposition is straightforward: it produces significantly smaller files than any other widely deployed codec at the same visual quality. Here is what that looks like in practice, measured at equivalent perceptual quality (VMAF 93, 1080p at 30fps):

AV1 delivers the same visual quality at 56% less bitrate than H.264. Compared to HEVC, the savings are around 30-35%. These numbers are consistent across a wide range of content types, though the exact improvement varies with content complexity — AV1’s advantage tends to be even larger on high-motion content like sports and gaming.

The downstream impact is substantial. Every byte you save on bitrate translates directly into reduced CDN costs, faster startup times for viewers on constrained networks, and less buffering. For a platform serving 1 million hours of video per month, switching from H.264 to AV1 could cut bandwidth costs by over 50%.

For a deeper comparison across all four major codecs, see our codec comparison guide.

#The encoding speed trade-off

Compression efficiency does not come free. AV1 requires significantly more compute to encode than older codecs:

# Approximate encoding speed on a modern 16-core CPU (1080p, software)
# H.264 (x264 slow):     ~120 fps — real-time on modest hardware
# HEVC  (x265 slow):     ~30  fps — 4x slower than H.264
# VP9   (libvpx-vp9):    ~15  fps — 8x slower than H.264
# AV1   (SVT-AV1 preset 5): ~10 fps — 12x slower than H.264
# AV1   (SVT-AV1 preset 8): ~25 fps — 5x slower than H.264

A few years ago, AV1 encoding at acceptable quality was 30-100x slower than H.264. The gap has narrowed considerably, primarily thanks to SVT-AV1 (Scalable Video Technology for AV1), which Intel and Netflix developed specifically to make AV1 encoding practical at scale.

SVT-AV1 achieves its speed improvements through several techniques:

Tiling and parallelisation. SVT-AV1 divides each frame into tiles that can be encoded independently across multiple CPU cores. Unlike libaom (the AV1 reference encoder), SVT-AV1 was designed from the ground up for multi-threaded execution, making it scale nearly linearly with core count on modern CPUs.

Preset flexibility. SVT-AV1 offers presets from 0 (highest quality, slowest) to 13 (fastest, lowest quality). Preset 5-6 offers a practical balance for production use: compression within 5-10% of the best possible, at speeds that are manageable for cloud encoding.

Continuous improvement. The SVT-AV1 team ships meaningful speed and quality improvements with each release. Between version 1.0 and the current 2.x series, encoding speed at equivalent quality has roughly doubled.

For cloud-based encoding, the speed trade-off is largely abstracted away. The encoding service handles parallelisation, hardware selection, and job scheduling. You pay a per-minute rate that reflects the additional compute, but you do not manage the infrastructure yourself.

#Hardware decode support

AV1 hardware decoding has reached a tipping point. The majority of devices sold in the last 2-3 years include AV1 decode support, which means AV1 playback is power-efficient and does not rely on CPU-based software decoding.

#Current hardware support matrix

#Browser support

#Smart TVs and streaming devices

The living room is where AV1 adoption matters most for streaming platforms. Most smart TVs manufactured since 2022 include AV1 hardware decode, including models from Samsung, LG, Sony, and Hisense. Streaming devices like Chromecast with Google TV, Amazon Fire TV Stick 4K (2022+), and Roku Ultra (2022+) all support AV1.

The notable gap is older devices. A 2019 smart TV or a first-generation Fire TV Stick will not have AV1 support. This is why a multi-codec strategy — serving AV1 to capable devices and falling back to H.264 or HEVC for older ones — remains the recommended approach.

#The royalty-free advantage

AV1 is developed by the Alliance for Open Media (AOM), a consortium that includes Google, Apple, Netflix, Amazon, Meta, Microsoft, Intel, AMD, NVIDIA, Samsung, and dozens of other companies. AV1 is released under a royalty-free patent license, which means you can encode, decode, and distribute AV1 content without paying licensing fees to anyone.

This stands in sharp contrast to HEVC, which is encumbered by three competing patent pools:

• MPEG-LA: The original HEVC patent pool
• HEVC Advance: A second pool with different patent holders and higher rates
• Velos Media: A third pool created by Nokia and others

The combined licensing cost of HEVC can reach $0.20 per device for hardware manufacturers and $0.05 per unit for content distributors, with annual caps that still run into millions of dollars for large-scale deployments. The licensing complexity alone has driven companies away from HEVC — it is widely cited as the primary reason that web browser vendors (other than Apple and Microsoft) never added native HEVC support to their desktop browsers.

AV1 eliminates this entire category of cost and legal complexity. At scale, the licensing savings alone can justify the higher encoding cost of AV1.

#Film Grain Synthesis

One of AV1’s most underappreciated features is Film Grain Synthesis (FGS) — a tool-level capability that Netflix productised and deployed broadly in July 2025.

Film grain is the visible texture in cinematic content shot on film or with intentional noise added for artistic effect. Traditional codecs treat film grain like any other visual detail and attempt to faithfully reproduce it, which consumes a disproportionate amount of bitrate. Grain is random and varies frame to frame, making it extremely expensive to encode.

AV1’s approach is different. The encoder analyses the grain characteristics of the source, strips the grain before encoding, and embeds grain parameters as metadata in the bitstream. The decoder then synthesises grain that statistically matches the original and applies it during playback.

The result: 15-30% bitrate reduction on cinematic content with heavy grain, with no perceptible difference in visual quality. The synthetic grain is generated from the same statistical model as the original, so it looks correct to human viewers even though it is not a frame-by-frame reproduction.

For platforms delivering cinematic content — films, prestige TV, documentaries shot on film — FGS can meaningfully reduce bandwidth costs without compromising the artistic intent of the content.

#When to start using AV1

If you are still evaluating whether AV1 belongs in your pipeline, here is a practical framework:

Use AV1 today if:

• Your audience primarily uses devices from 2021 or later
• Your content is consumed in web browsers (Chrome, Firefox, Edge, Safari 17+)
• You are delivering to smart TVs manufactured since 2022
• CDN costs are a significant portion of your video infrastructure spend
• You are already using adaptive streaming (HLS/DASH) with multi-codec manifests

Wait on AV1 if:

• Your audience includes significant numbers of pre-2020 devices
• You are targeting embedded systems or IoT devices with fixed hardware
• Your encoding volume is low enough that CDN savings would not offset the encoding premium
• You are in a live streaming context where encoding latency is critical (AV1 live encoding is still maturing)

The recommended approach for most platforms is a multi-codec ladder: encode every piece of content in both H.264 and AV1, and let the player select AV1 on capable devices. Viewers on modern hardware get smaller files and faster startup; viewers on older hardware get the H.264 fallback. You serve every device while capturing AV1’s efficiency gains where they are available.

#AV1 with Transcodely

Transcodely supports AV1 encoding using SVT-AV1, with all three quality tiers (Economy, Standard, Premium). The per-minute rate for AV1 at 1080p is EUR 0.04 — higher than H.264’s EUR 0.01, reflecting the additional compute required.

But the total cost picture often favours AV1. Here is why: encoding is typically 10-20% of your total video delivery cost. CDN bandwidth and storage are the dominant expenses. If AV1 cuts your file sizes by 50%, the savings on delivery can far outweigh the encoding premium.

#Single-codec AV1 output

{
  "source_url": "https://storage.example.com/raw/documentary.mp4",
  "outputs": [
    {
      "format": "hls",
      "codec": "av1",
      "quality": "standard",
      "resolutions": ["2160p", "1080p", "720p", "480p"]
    }
  ],
  "storage": {
    "provider": "s3",
    "bucket": "my-cdn-origin",
    "path": "encoded/{job_id}/"
  },
  "webhook_url": "https://api.example.com/hooks/transcode"
}

#Multi-codec adaptive streaming (H.264 + AV1)

{
  "source_url": "https://storage.example.com/raw/documentary.mp4",
  "outputs": [
    {
      "format": "adaptive",
      "codecs": ["h264", "av1"],
      "quality": "standard",
      "resolutions": ["1080p", "720p", "480p"],
      "thumbnails": {
        "interval_seconds": 10,
        "width": 320
      }
    }
  ],
  "storage": {
    "provider": "s3",
    "bucket": "my-cdn-origin",
    "path": "encoded/{job_id}/"
  },
  "webhook_url": "https://api.example.com/hooks/transcode"
}

The adaptive format generates CMAF segments with both HLS and DASH manifests, serving H.264 renditions to older devices and AV1 to modern ones — all from the same set of segment files.

#The cost math

For a 10-minute video at 1080p:

The AV1 encoding premium (EUR 0.30 extra) pays for itself after approximately 43 views in CDN savings. For any content that gets more than a few dozen views, AV1 is the more cost-effective choice.

For a detailed breakdown of how to optimise your total video costs beyond just encoding, see our guide to reducing video encoding costs.

#What is next: AV2

The Alliance for Open Media is actively developing AV2, the successor to AV1. While the specification is not yet finalised, early results indicate another 30-40% compression improvement over AV1, continuing the generational trend.

AV2 hardware decode support is expected to appear in consumer devices starting in late 2026 and 2027. Like AV1, it will be royalty-free.

The practical implication: if you build your pipeline to support multiple codecs today, adding AV2 when it becomes viable will be a configuration change rather than an architecture overhaul. The multi-codec adaptive streaming approach that serves AV1 to modern devices and H.264 to older ones will simply gain a third tier — AV2 for the newest devices, AV1 for the middle generation, and H.264 for everything else.

The codec ladder gets longer, but the principle remains the same: encode once per codec, serve the most efficient option to each viewer, and let the player negotiate the best available quality.

AV1 is no longer the future of video compression. It is the present. The question is not whether to adopt it, but how quickly your pipeline can start capturing its benefits.