Fact-checked by Grok 2 weeks ago

WebP

WebP is a modern raster image file format developed by Google, utilizing both lossy and lossless compression algorithms derived from the VP8 video codec to enable smaller file sizes while maintaining high visual quality for web images. It supports transparency (alpha channel) in both compression modes and animated sequences, serving as an efficient alternative to established formats like JPEG for photographs, PNG for graphics with transparency, and GIF for simple animations. Introduced on September 30, 2010, as an open-source standard, WebP aims to accelerate web page loading by reducing image data, which accounts for a significant portion of web traffic. Key advantages of WebP include substantial reductions: lossy WebP images are typically 25–34% smaller than equivalent JPEGs at comparable levels, while lossless WebP files are about 26% smaller than PNGs. was added on November 18, 2011, enhancing its versatility for scenarios requiring exact image reproduction, such as diagrams or icons, and enabling alpha that results in files up to three times smaller than PNG equivalents in lossy mode. Animated WebP, introduced later to support frame-by-frame sequences with optional lossy or lossless encoding and , offers significantly reduced sizes compared to GIF and APNG, promoting faster rendering on devices. WebP has gained broad adoption, with native support in major browsers including (since version 17 for lossy), Mozilla Firefox, Apple Safari, , and , as well as in operating systems like and . The format's , libwebp, provides encoding and decoding tools (cwebp and dwebp) under a BSD license, facilitating integration into web servers, systems, and image editing software. Despite its benefits, WebP's rollout faced initial challenges due to limited early browser compatibility, but by 2025, it has become a standard recommendation for optimizing .

History

Development and Initial Release

Google developed WebP as an initiative to enhance by significantly reducing the size of image files, addressing the fact that images accounted for approximately 65% of bytes at the time. The format was publicly announced on , 2010, as a and open-source format under a BSD license, leveraging technology from the video codec that Google had open-sourced in May 2010 as part of the multimedia project. The core motivations behind WebP included achieving superior compression efficiency over established formats like for lossy encoding and for lossless encoding, with planned extensions for alpha transparency and capabilities to serve as a modern alternative to . Initial development focused on adapting VP8's intra-frame video compression techniques to still images within a lightweight RIFF-based , aiming to enable faster page loads and lower bandwidth usage without sacrificing visual quality. Google released a developer preview alongside the announcement, featuring the libwebp reference library for encoding and decoding, along with command-line tools for converting existing images to WebP. Integration with the rendering engine was planned from the outset, with a patch developed for to enable native support in an upcoming release of , beginning with version 9 in early 2011. Early evaluations by highlighted WebP's potential impact, claiming an average 39% reduction in file size compared to across a dataset of over one million web images, while maintaining equivalent perceptual quality as measured by standard metrics. These benchmarks underscored WebP's role in 's broader push for an open media ecosystem.

Subsequent Updates and Milestones

Lossless compression for WebP was announced on November 18, 2011, with initial integration into Google Chrome 17. This update built on the format's foundational lossy compression by adding support for lossless modes, enabling smaller file sizes for graphics without quality degradation in scenarios requiring exact reproduction. The Chrome integration marked a key step toward practical deployment, allowing developers to test WebP on live web pages without plugins. libwebp version 0.2.0, released in August 2012, enabled lossless support by default. In 2012, adoption accelerated with adding native WebP support in version 12.1, expanding compatibility beyond and encouraging broader experimentation. In October 2011, Google released the first public samples of animated WebP files, demonstrating the format's potential for replacing GIFs with more efficient, high-quality animations featuring 24-bit . A major milestone arrived in 2018 with the stable release of libwebp 1.0 on April 21, which optimized encoding speeds by up to 30% for lossy modes and improved overall quality metrics through refined chroma handling and reduced artifacts. These enhancements made WebP more viable for high-volume encoding workflows, such as content delivery networks, without compromising on efficiency. Between 2020 and 2023, WebP gained deeper integration across mobile ecosystems, including native support in (API level 30), which streamlined decoding and encoding for apps and system-level image handling. This period also saw expanded use in systems and libraries, driving practical adoption in mobile-first . By 2024 and into 2025, WebP became the default format in key services like Search and . The announcement of WebP 2 in June 2021 served as a forward-looking , aiming to achieve compression parity with emerging formats like while maintaining . An important enabler for widespread adoption was the resolution of patent concerns in , when reached a licensing agreement with and 11 patent holders covering techniques essential to WebP, confirming its royalty-free status and alleviating legal uncertainties.

Technology

Core Features and Container Format

WebP is a raster graphics format developed by that employs the (RIFF) as its container structure, with the file beginning with a 'WEBP' chunk to denote the format. This chunk-based RIFF design allows for modular organization of data, enabling the inclusion of various components such as image streams, , and extensions within a single file up to approximately 4 GiB in size. The format's flexibility supports both still images and animations, making it suitable for web delivery. A key strength of WebP lies in its unified support for both lossy and lossless compression modes, alongside an 8-bit alpha channel that facilitates transparency in RGBA color spaces with 24-bit RGB depth. Lossy images use the VP8 chunk for data storage, while lossless ones utilize the VP8L chunk; the VP8X chunk signals extended capabilities like alpha blending or animation when present. This integration allows WebP to handle photographic content efficiently in lossy mode and graphics with sharp edges or transparency in lossless mode, all within the same container. WebP accommodates embedded metadata through dedicated chunks, including for exchangeable image file format data, for descriptions, and profiles for , all nested under the VP8X extension for enhanced interoperability. The standard file extension is .webp, and the registered type is image/webp, ensuring broad recognition across web platforms and applications. At its core, the WebP encoding pipeline begins by partitioning the input image into 16x16 macroblocks, which are subsequently refined through intra-prediction to exploit spatial redundancies and to compress the residual data efficiently. This high-level process underpins the format's performance, yielding typical file size reductions of 25-34% over equivalent-quality images in lossy mode and approximately 26% over in lossless mode. When transparency is involved, lossy WebP with lossless alpha can achieve 60-70% smaller sizes than transparent files.

Lossy Compression

WebP's mechanism derives from the intra-frame encoding of the video codec, where static s are processed as individual key frames in a video , utilizing block-based to exploit spatial redundancies within the . The encoding begins by partitioning the into 16×16 macroblocks, which are subdivided into smaller units: 4×4 blocks for (brightness) details, 16×16 blocks for broader luma , and 8×8 blocks for (color) components. This approach allows for targeted of and separately, enhancing overall efficiency. Spatial prediction forms a core technique, employing ten distinct modes per 4×4 luma block to estimate values from already decoded neighboring pixels. These modes reduce the that needs to be encoded by capturing local correlations effectively. The residuals undergo a (DCT) on 4×4 blocks, which converts spatial into frequency-domain coefficients, concentrating most energy in low-frequency components and producing many zeros for subsequent . This is followed by arithmetic , which adaptively assigns shorter codes to frequent symbols, outperforming traditional in efficiency. Quantization introduces the irreversible loss, scaling the DCT coefficients by a quantization parameter (QP) to discard less perceptible high-frequency details, controlled via an adjustable quality factor from 0 (maximum compression, lowest quality) to 100 (minimal loss, highest quality). To further optimize color encoding, WebP uses YUV 4:2:0 chroma subsampling, halving the horizontal and vertical resolution of chroma channels relative to luma, as human vision is less sensitive to color variations than brightness. Compared to JPEG, WebP's lossy mode delivers superior quality at equivalent low bitrates, thanks to advanced spatial prediction that better captures edges and textures, combined with in-loop deblocking and deringing filters that mitigate artifacts like blocking. These enhancements, along with adaptive per-block quantization, enable 25–34% smaller file sizes at matched perceptual quality metrics such as SSIM. Encoding remains computationally intensive, typically 10–20 times slower than JPEG due to the complex prediction and transform processes, but decoding is faster, benefiting web rendering.

Lossless Compression

WebP's lossless compression mode, denoted as VP8L, enables the exact reproduction of original data without any quality loss, making it a suitable replacement for formats like . This mode employs an entropy-coded spatial prediction scheme that exploits correlations between neighboring pixels to reduce redundancy. Specifically, it uses 14 different predictors applied in scan-line order, where each pixel's value is estimated based on its left (L), top (T), top-left (TL), and top-right (TR) neighbors, with the residual encoded to minimize bits. Additionally, backward references inspired by LZ77 allow referencing previous segments, using distance and length codes that prioritize short distances within a 120-pixel neighborhood for efficiency. A color further optimizes storage by indexing up to 2^11 recently used colors via a , reducing the encoding of repeated color values. The compression process begins with a series of reversible transforms to decorrelate the image data. The subtract-green transform removes the green channel's influence from red and blue channels by subtracting green values, which is particularly effective for natural images where green dominates. For images with limited colors, a palette transform indexes pixels to a table of up to 256 entries, bundling multiple pixels per code when the palette size is small (≤16 colors). The predictor transform divides the image into blocks and selects one of the 14 predictors per block using the green channel value as an index. A color transform then applies a reversible decorrelation similar to the YCoCg-R space, preserving the green channel while adjusting red and blue based on predefined deltas (e.g., red transformed as R - ((B + G) >> 1), with inverse operations ensuring bit-exact reversibility). Following these transforms, entropy coding uses spatially variant Huffman tables—up to five per pixel group for green, red, blue, alpha, and distances—with prefix codes that adapt to local statistics, typically achieving 14-19 bits per pixel depending on image complexity. WebP supports lossless transparency through direct encoding of the alpha as a separate plane, integrated into the ARGB without any or lossy approximations, ensuring precise preservation of semi-transparent regions. This approach contrasts with formats that might compress alpha indirectly, allowing VP8L to maintain full for images with varying opacity. In terms of efficiency, VP8L achieves compressed file sizes approximately 26% smaller than equivalent PNGs on average, as demonstrated across diverse web image corpora, owing to the combined effects of spatial , backward references, and color caching that better exploit image redundancies than PNG's DEFLATE-based method. The can be quantified as \text{Compression ratio} = \frac{\text{original size} - \text{compressed size}}{\text{original size}}, where VP8L typically yields ratios 22-30% higher than optimized PNG variants through these integrated techniques, with studies showing 23% improvement over size-optimized PNGs like ZopfliPNG for translucent images. While VP8L encoding is generally slower than PNG due to the iterative optimization of transforms and Huffman tables for maximum density, decoding performance is comparable or faster, averaging around 0.003 seconds per image versus 0.005 seconds for PNG on typical web content, facilitating efficient rendering in resource-constrained environments.

Animation Support

Animated WebP extends the format's capabilities to support animated images through the (RIFF) container, utilizing specific chunks to define sequences and playback parameters. The core structure includes an 'ANIM' chunk for global animation settings, such as count and background color, followed by multiple 'ANMF' chunks, each representing an individual in display order. Each 'ANMF' chunk encapsulates , including position (X and Y offsets, 24-bit values scaled by 2), dimensions (width and height minus one, 24-bit), and duration in milliseconds (24-bit value). This allows for sequences of full or partial , enabling efficient incremental updates through , where subsequent reference and modify only changed regions rather than the entire . Frames in animated WebP can employ via the codec, via VP8L, or a mixed approach combining VP8 with an 'ALPH' chunk for . Delta frames leverage inter-frame , particularly in lossy mode, to reduce redundancy by encoding differences from previous frames, similar to video codecs. This , along with support for partial frame updates, contributes to significant efficiency gains: animated WebP files are typically 64% smaller than equivalent GIFs when using , based on conversions of a corpus of approximately 7,000 GIFs. Playback control mirrors aspects of GIF, with looping defined by the 'ANIM' chunk's 16-bit loop count (0 for infinite repetition) and a specified background color applied during disposal. Disposal methods are limited to two options per frame: none (retaining the frame as is for the next) or dispose to background (clearing the frame rectangle to the background color). Blending modes further refine rendering: alpha-blending (0) composites the frame over the previous one using transparency, while replace (1) discards the prior frame content before drawing the new one. Frame durations are precisely controlled in milliseconds, allowing smooth animations with fine-grained timing. The libwebp library provides tools for encoding animated WebP, notably img2webp, which converts sequences of input images (e.g., or ) into animated files. It supports flags such as -loop <count> to set the number of repetitions (default infinite) and -d <ms> to specify per-frame delay (default 100 ms), alongside options for lossless/lossy modes and keyframe intervals to optimize compression and seeking.

Comparisons with Other Formats

Versus JPEG and PNG

WebP's lossy compression offers significant advantages over for photographic images, achieving file sizes that are 25-34% smaller while maintaining equivalent visual quality as measured by the Structural Similarity Index (SSIM). This efficiency stems from WebP's use of VP8-based encoding, which employs and adaptive quantization to distribute bits more evenly across the image, reducing visible artifacts compared to 's discrete cosine transform (DCT) method. In particular, WebP minimizes blocky artifacts around edges and high-contrast areas that are common in at similar compression levels, providing sharper details without the characteristic 8x8 pixel blocking. benchmarks from 2010 onward demonstrate that WebP encoded at 80% quality often matches or exceeds the perceptual quality of at 90% quality across diverse datasets like Kodak and Tecnick, with SSIM scores typically above 0.95 for both but WebP requiring fewer bits per pixel. For lossless compression, WebP provides a compelling alternative to , particularly for , , and images requiring , with file sizes approximately 26% smaller on average. Evaluations confirm WebP outperforms standard PNG libraries by 42% and optimized variants like ZopfliPNG by 23% in terms of compression density, while preserving perfect fidelity (SSIM of 1.0). Unlike , where via alpha channels is optional and can increase file sizes substantially for opaque images, WebP integrates alpha support natively in its for both lossy and lossless modes, enabling efficient handling of semi-transparent elements without additional overhead. This makes WebP preferable for web where PNG's palette-based optimization excels for simple colors but falls short in spatial prediction for complex scenes. In practical use cases, WebP lossy is favored over for web photographs due to faster loading times and reduced , especially on mobile devices, while its lossless mode suits PNG's domain of icons and diagrams by combining smaller sizes with built-in . Additionally, WebP's royalty-free status under a BSD license addresses historical licensing concerns associated with 's , promoting broader adoption without legal barriers.

Versus GIF and AVIF

WebP animations provide significant advantages over , primarily through superior and efficiency. Unlike , which is limited to an 8-bit palette supporting only 256 colors, WebP animations utilize 24-bit RGB , enabling millions of colors without the need for dithering artifacts that often degrade quality in complex visuals. This full-color support makes WebP ideal for vibrant, detailed animations where 's palette restrictions lead to banding or posterization. Additionally, Google's benchmarks on a corpus of approximately 7,000 animated s demonstrate that lossy WebP conversions achieve a 64% reduction compared to the originals, resulting in files roughly three times smaller on average, while lossless WebP yields a 19% reduction. These efficiencies stem from WebP's advanced algorithms, allowing for high-quality animations with reduced demands suitable for delivery. In comparison to , a newer format based on the video codec, WebP offers a balance of performance and compatibility, though AVIF excels in certain areas. An experimental codec known as WebP 2, under development, aims for compression efficiencies similar to AVIF with about 30% better than WebP 1.x and support for 10-bit . However, current implementations of WebP (version 1.x) provide faster encoding and decoding speeds, making it more practical for applications where AVIF's AV1 foundation results in slower processing—often 2-4 times longer decode times. AVIF demonstrates superior , producing files 20-30% smaller than equivalent WebP images at matched quality levels, particularly for photographic content, and it outperforms in scenarios with up to 12-bit versus WebP's standard 8-bit. For animations, AVIF's support remains incomplete in 2025, with browser implementations like requiring experimental flags and offering only partial sequence handling, limiting its reliability compared to WebP's mature animation framework. Key trade-offs highlight WebP's established maturity against AVIF's open, AV1 foundation, which promotes broader long-term adoption without licensing concerns—though both formats are royalty-free overall. WebP's longer development history ensures better legacy compatibility across older devices and software, reducing fallback needs in mixed environments. As of November 2025, WebP sees higher usage with approximately 18.1% of websites implementing it, including many top sites, compared to AVIF's 1.0% overall penetration, reflecting WebP's edge in practical deployment despite AVIF's growing browser support at around 93%.

Support and Adoption

Web Browser Support

WebP has achieved widespread native support across major web browsers by 2025, enabling efficient rendering of its lossy, lossless, animated, and transparent variants without requiring plugins. Google Chrome introduced initial support for basic lossy WebP images in version 17, released in October 2011. Full support, including lossless compression, alpha transparency, and animation, arrived in Chrome 32 in January 2014. With Chrome holding approximately 66% of the global browser market share as of Q3 2025, nearly 99% of Chrome users benefit from complete WebP compatibility. Mozilla added full WebP support in version 65, released in January 2019, covering all features including animation and transparency; prior versions offered no support. As commands about 3.8% of the market as of Q3 2025, this equates to roughly 95% coverage among its user base. Apple implemented WebP support starting with version 14 in September 2020, requiring (11) or later for desktop and for mobile devices. By 2025, with iOS adoption exceeding 90% on compatible devices, users largely experience full feature support. Microsoft Edge gained full WebP support in version 18, released in 2018, following its transition to the engine; the legacy never supported WebP. provided partial support from version 11.5 in 2011, with full compatibility—including animation—arriving in version 19 in 2013. As of 2025, WebP enjoys approximately 97% global browser support, leaving less than 1% of users—primarily on outdated versions—without native decoding. Developers commonly employ fallback strategies, such as the HTML <picture> element to serve alternative formats like or , or polyfills for legacy browsers.
BrowserInitial SupportFull Support (incl. Animation/Transparency)Notes
v17 (2011, lossy only)v32 (2014)Dominant ensures broad coverage.
None before v65v65 (2019)Comprehensive from launch.
None before v14v14 (2020, OS-dependent)Requires macOS 11+ or +.
None before v18v18 (2018)IE11 unsupported.
v11.5 (2011, partial)v19 (2013)Aligns with timeline.

Graphics Software Support

Various graphics software applications have integrated support for creating, editing, and converting WebP files, often leveraging the open-source libwebp library developed by for encoding and decoding capabilities. This library underpins WebP functionality in many tools, enabling both and as well as handling where supported. Command-line utilities like cwebp, part of the libwebp suite, are commonly used for batch conversions and scripting in professional workflows. Adobe Photoshop provides WebP support through the WebPShop , which has been available since 2011 for opening, editing, and saving WebP images, including animations. Native integration was added in Photoshop 23.2, released in 2022, allowing direct import and export without plugins, though some advanced features like encoding previews remain plugin-dependent. Import operations in Photoshop rely on libwebp for decoding. The image editor has offered native WebP support since version 2.10, released in May 2018, including full encoding and decoding for both static and animated files, along with and metadata preservation. supports WebP via a dedicated developed around 2015, which enables loading and saving of WebP images. This plugin became bundled natively starting with version 4.2.5, released on October 1, 2019, providing seamless integration for users. IrfanView and , popular for image viewing and , have provided full WebP read/write support since version 4.40 (2012) via plugins, facilitating efficient conversion of large image sets. Both tools excel in batch operations, allowing users to convert formats like or to WebP while preserving quality settings. Affinity Photo introduced WebP export support in version 2.0 (November 2022), with subsequent updates enhancing compatibility for professional editing tasks. By 2024, version 2.4 further improved export options, including better handling of lossless modes for web-optimized workflows.

Other Applications and Libraries

WebP has been integrated into various operating systems for native or extended viewing capabilities. On and later versions, WebP images can be viewed using the built-in Photos app and thumbnails following the installation of the official WebP Image Extension, which has been available since 2018 and provides support for lossy, lossless, and animated formats. For macOS, WebP support is available through WebKit-based applications since Safari 14, released in 2020, enabling decoding in compatible environments. On Linux distributions, WebP is supported via tools like , which includes encoding and decoding options for the format in its core functionality. In mobile operating systems, WebP decoding has been natively available on since version 4.0 ( level 14, released in 2011), supporting , with lossless and transparency features added in Android 4.3 ( level 18). Encoding capabilities for WebP are facilitated through Android's .compress method since level 14 for lossy images and level 17 for lossless, though advanced features like animated WebP decoding were expanded in Android 9 and later. On , WebP support is provided via since (2020), allowing decoding in and other WebKit-dependent apps. Developer libraries play a key role in WebP integration beyond core OS features. Google's libwebp, a C-based reference for encoding and decoding WebP images, reached 1.6.0 in June 2025, incorporating improvements such as optimized x86 instructions (AVX2 and ) for faster processing. Wrappers like for provide high- WebP encoding and decoding, supporting conversions from formats like , , and while handling lossless, lossy, and animated variants. Similarly, Python's supports WebP operations, including opening, saving, and converting images, provided the underlying libwebp is installed on the system. Several applications leverage WebP for efficient image handling. has supported uploading, displaying, and using WebP images since version 5.8 (released in July 2021), treating them equivalently to and files in media libraries and themes. Messaging platforms like Telegram utilize WebP exclusively for static stickers, requiring images to be in WebP format with specific dimensions (e.g., 512 pixels on one axis) and transparency support. added support for animated WebP images in March 2025, enhancing media uploads and displays across its infrastructure for reduced file sizes compared to . Server-side and content management adoption of WebP continues to grow through tools like version 7.1 and later, which offer robust WebP encoding options including quality settings, lossless modes, and animation handling for in web environments.

Criticisms and Limitations

Compatibility Challenges

One significant compatibility challenge for WebP arises from its lack of support in legacy web browsers. does not support WebP, and versions of prior to 14.0 also lack native decoding capabilities, requiring macOS 11 or later for implementation in Safari 14 through 15.6. Although these browsers account for less than 1% of global usage in 2025, they remain prevalent in enterprise environments where updating systems is often restricted by security policies or legacy software dependencies. In email applications, WebP adoption is uneven, complicating its use in communications. Outlook's desktop client does not provide native support for WebP images, necessitating fallbacks to formats like or to ensure display. has supported WebP since 2011, but it often automatically converts WebP files to for broader compatibility, potentially losing lossless details or transparency. Hardware-level support for WebP decoding is limited, particularly on older graphics processing units (GPUs) and peripheral devices. WebP relies primarily on software decoding in most implementations, which can increase on devices with limited CPU resources. Similarly, many printers and ignore WebP files due to their focus on established formats like and , requiring manual conversion before use in printing or scanning workflows. To address these gaps, developers often employ conversion tools and strategies that add overhead. Google's official cwebp command-line tool and libraries like libwebp enable batch conversion of WebP to compatible formats for non-supporting platforms, while online services such as CloudConvert provide quick transformations without installation. A common web practice is double-serving images—delivering WebP to supporting clients via the <picture> element and fallbacks to or otherwise—which ensures but doubles and demands on servers. Criticisms of WebP's rollout highlight how Google's promotion through tools like PageSpeed Insights has accelerated browser adoption but contributed to a perception of fragmented web standards. Early resistance from software vendors, including delayed in until , slowed ecosystem integration, forcing developers to maintain multiple format pipelines and exacerbating interoperability issues in diverse environments.

Performance and Quality Issues

In lossy mode, WebP images at quality settings below 50 often display prominent ringing and blocking artifacts, manifesting as halo-like distortions around edges and grid-like patterns in uniform areas. These issues are exacerbated in smooth gradients, where WebP performs worse than JPEG, producing posterization effects such as unintended magenta or green rings in scanned film or portraits. WebP encoding is significantly slower than that of JPEG or PNG, typically 5-10 times longer for lossy compression due to its advanced prediction and transformation techniques, making it resource-intensive for batch processing or real-time applications. While decoding is relatively efficient—about 1.5 times slower than JPEG—it remains CPU-demanding, particularly on mobile devices with limited processing power. Objective quality metrics reveal WebP's limitations compared to newer formats; for instance, achieves higher PSNR and SSIM scores at equivalent bitrates, indicating better preservation of structural details and reduced distortion. As of 2025, WebP remains limited to depth per channel without native support for (), leading to color banding in images with subtle gradients or high dynamic range, unlike which supports up to 12-bit and . Real-world bandwidth savings with WebP are more modest than initially claimed, averaging 15-25% file size reduction over in practical web scenarios, depending on image content like photographs versus graphics, rather than the advertised 30% or more. Critics argue that has overhyped WebP's advantages, with its quality gains often marginal and artifacts making preferable for crisp, line-art graphics where lossless fidelity without banding is essential.

Security Vulnerabilities

Known Issues and CVEs

One of the most critical vulnerabilities in libwebp, the for WebP, is CVE-2023-4863, a heap buffer overflow in the lossless decoder (VP8L) that enables out-of-bounds memory writes. This flaw, with a CVSS score of 8.8, stems from inadequate validation during Huffman code table allocation and access, allowing attackers to craft malformed WebP files that overflow allocated buffers. It affects libwebp versions prior to 1.3.2 and was actively exploited in the wild, impacting applications like before version 116.0.5845.187. Earlier vulnerabilities include multiple integer overflows in libwebp, such as those documented in CVE-2016-9085, which allow unspecified impacts through crafted inputs during decoding. These flaws arise from improper handling of integer values in decoding routines, leading to buffer overflows similar to later issues. In 2025, Google Project Zero detailed an exploit chain (BLASTPASS) that leveraged a WebP vulnerability in for zero-click attacks via , highlighting ongoing risks in image parsing. Common root causes across these vulnerabilities involve integer overflows during chunk parsing in WebP files, such as in ParseOptionalChunks, where unvalidated sizes lead to incorrect memory allocations and subsequent remote code execution potential. For instance, oversized frame or chunk dimensions in animated WebP files can cause stride calculation overflows, enabling corruption. These issues broadly affect decoders in web browsers like and Electron-based applications, as well as other software embedding libwebp.

Exploitation Cases and Mitigations

In 2023, a zero-day vulnerability designated CVE-2023-4863 was actively exploited in Google Chrome versions prior to 116.0.5845.187, enabling remote attackers to perform out-of-bounds memory writes via specially crafted WebP images processed by the libwebp library. This flaw was patched in Chrome 116.0.5845.187 on September 12, 2023, following reports of in-the-wild exploitation. A notable exploitation case involved NSO Group's Pegasus spyware in the BLASTPASS attack chain, which targeted iOS devices via iMessage attachments containing malicious WebP images in lossless format. Delivered as PKPass files, these exploits leveraged the WebP vulnerability within Apple's ImageIO framework to achieve zero-click remote code execution, bypassing the BlastDoor sandbox and enabling surveillance on iOS versions up to 16.6. The attack was captured in the wild in 2023, with detailed analysis released by Google Project Zero in March 2025 confirming the role of Huffman table overflows in the VP8L chunk. Additional exploitation vectors emerged in Electron-based applications, where the vulnerability allowed remote code execution through HTML tags rendering untrusted WebP content, affecting apps like and prior to library updates. On Android, the flaw impacted image-processing applications, including gallery apps that decode WebP files, potentially leading to heap overflows when handling malicious media from unverified sources in 2023. To mitigate these risks, released libwebp version 1.3.2 on September 13, 2023, addressing the core in the WebP codec. Browser vendors enhanced protections through (ASLR) and continuous fuzzing of image decoders, while site isolation and renderer sandboxes in Chromium-based browsers confined potential remote code execution to isolated processes, reducing system-wide impact. Security recommendations emphasize server-side validation of WebP images to detect malformed files before , alongside mandatory updates to libwebp 1.3.2 or later in all dependent applications. Web developers are advised to sanitize untrusted uploads and avoid direct rendering of external WebP content without verification, as highlighted in post-incident guidance from 2023. In response, the Chromium project intensified fuzzing efforts on libwebp and related codecs to proactively identify similar flaws, contributing to broader ecosystem hardening against image-based attacks.

WebP 2

Development and Goals

Google announced the development of WebP 2 in late 2020 as a successor to the original WebP format, aiming to enhance its capabilities in response to evolving image compression standards. This initiative was driven by the need to address limitations in WebP 1.0, particularly its slower encoding speeds compared to decoding, and to position it competitively against newer formats like AVIF, which leverages the AV1 video codec for superior efficiency. The project maintains WebP's royalty-free status under the BSD license, ensuring broad accessibility for web developers and applications. The primary goals of WebP 2 include achieving 20-30% better compression efficiency over WebP 1.0 for lossy images, while preserving or improving support for , , , and incremental decoding. It also introduces support for (HDR) imaging via a 10-bit architecture compatible with , enabling richer visuals on modern displays. These enhancements target reduced file sizes for and use cases, ultimately improving load times and user experience without sacrificing quality. As of 2025, WebP 2 remains in an experimental prototype phase, with the libwebp2 library serving as its integrated into 's for testing. No full public release as a has occurred, though alpha-level tools and binaries are available through the project's for developers to experiment with techniques. The development, led by engineers focused on performance optimization, emphasizes to ensure existing WebP 1.0 files can be processed seamlessly.

Planned Features and Status

WebP 2 is an experimental image codec developed by , building on the original WebP format to explore advancements in compression and related capabilities. It focuses primarily on enhancing lossy and , achieving approximately 30% better in lossy modes compared to WebP 1.0, as demonstrated in internal benchmarks. Key experimental features include support for 10-bit HDR encoding, improved methods, animation handling with incremental decoding, and full multi-threading for both encoding and decoding processes to optimize on modern hardware. These enhancements aim to better handle web-typical content such as photographs, graphics with transparency, and short animations, while maintaining compatibility with existing WebP tools where possible. As of November 2025, the project remains active in 's repository with ongoing maintenance activity. However, has explicitly stated that WebP 2 will not be released as a production image format; instead, it serves as a to test ideas that could inform future iterations of libwebp or other technologies. No public , trials, or efforts, such as through the W3C, are planned.

References

  1. [1]
    An image format for the Web | WebP - Google for Developers
    Aug 7, 2025 · WebP is a modern image format that provides superior lossless and lossy compression for images on the web. Using WebP, webmasters and web ...WebP Gallery · WebP Image Galleries · WebP Container Specification · Guides
  2. [2]
    WebP Container Specification - Google for Developers
    Aug 7, 2025 · WebP is an image format that uses either (i) the VP8 key frame encoding to compress image data in a lossy way or (ii) the WebP lossless encoding.
  3. [3]
    WebP, a new image format for the Web - Google for Developers Blog
    Sep 30, 2010 · We launched the Page Speed Firefox extension to evaluate the performance of web pages and to get suggestions on how to improve them, we ...Missing: first | Show results with:first
  4. [4]
    Lossless and transparency encoding in WebP
    Nov 18, 2011 · WebP lossy with alpha can be used to create transparent images that have minimal visual degradation, yet are much smaller in file size.
  5. [5]
    Frequently Asked Questions | WebP - Google for Developers
    Oct 7, 2025 · The WebP format essentially aims at creating smaller, better looking images that can help make the web faster. Which web browsers natively ...
  6. [6]
    Getting Started | WebP - Google for Developers
    Aug 7, 2025 · WebP is a modern image format supported by various software and browsers like Google Chrome and Opera. · The libwebp codec provides tools like ...
  7. [7]
    WebM and VP8 land in Chromium
    Wednesday, May 19, 2010. A developer preview of WebM, a high-quality, open, freely implementable, and web-optimized video format was announced today.
  8. [8]
    WebP image format | Can I use... Support tables for HTML5, CSS3, etc
    Image format (based on the VP8 video format) that supports lossy and lossless compression, as well as animation and alpha transparency.
  9. [9]
    New WebP tools released! - Google Groups
    Dear all,. as of a few hours ago we released to the open source repository the biggest update to WebP to date.
  10. [10]
    WebP - Wikipedia
    WebP is a raster graphics file format developed by Google intended as a replacement for JPEG, PNG, and GIF file formats. It supports both lossy and lossless ...History · Technology · Support · Disadvantages and criticism
  11. [11]
    WebP Support - KeyCDN
    Jul 16, 2020 · Opera has officially had WebP support since Opera 12.1 (originally released in November, 2012), and partial support since Opera 11.5. ... support ...
  12. [12]
    The Rise of WebP - WHITE64
    Apr 29, 2022 · In October of 2011, Google improved the WebP format by adding the ability to support frame-by-frame animation. Previously only the GIF format ...
  13. [13]
    Create WebP images | Android Studio
    Apr 12, 2023 · Android Studio can convert PNG, JPG, BMP, or static GIF images to WebP format. You can convert individual images or folders of images.Missing: 2020-2023 | Show results with:2020-2023
  14. [14]
    Usage statistics of WebP for websites - W3Techs
    These diagrams show the usage statistics of WebP as image file format on the web. See technologies overview for explanations on the methodologies used in the ...
  15. [15]
    Google and MPEG LA settle long-running VP8/H.264 patent dispute
    Mar 7, 2013 · That's the case with the long-running dispute over Google's VP8 codec, which was settled today after two years of legal machinations. Back in ...Missing: issues | Show results with:issues
  16. [16]
    RFC 9649 - WebP Image Format - IETF Datatracker
    Nov 18, 2024 · WebP is an image file format based on the Resource Interchange File Format (RIFF) [RIFF-spec] (Section 2) that supports lossless and lossy compression as well ...
  17. [17]
    Compression Techniques | WebP - Google for Developers
    Aug 7, 2025 · WebP is a new image format developed by Google and supported in Chrome, Opera and Android that is optimized to enable faster and smaller images on the Web.
  18. [18]
    WebP Compression Study - Google for Developers
    Aug 7, 2025 · The study evaluated WebP compression in comparison to JPEG. We observed that the average WebP file size is 25%-34% smaller compared to JPEG file size at ...<|separator|>
  19. [19]
    None
    Nothing is retrieved...<|separator|>
  20. [20]
    Specification for WebP Lossless Bitstream | Google for Developers
    Aug 7, 2025 · This document describes the compressed data representation of a WebP lossless image. It is intended as a detailed reference for the WebP lossless encoder and ...
  21. [21]
  22. [22]
    RFC 9649: WebP Image Format
    2.7.1.1. Animation · The frame disposal only applies to the frame rectangle, that is, the rectangle defined by Frame X, Frame Y, frame width, and frame height.
  23. [23]
    WebP tools
    ### Tools for Encoding Animated WebP
  24. [24]
    aa3651e5c43a2c22d75ba500e0...
    Jul 24, 2025 · The WebP 2 experimental codec is mostly pushing the features of WebP further in terms of compression efficiency. The new features (like 10b HDR ...
  25. [25]
    WebP vs AVIF - Which is better in 2025? | SpeedVitals Blog
    Aug 20, 2024 · AVIF is better than WebP in terms of Compression level and quality. However, WebP takes less time to decode and it is more widely supported.
  26. [26]
    AVIF vs WebP: Which Modern Image Format Is Better? - ShortPixel
    WebP uses V8 compression technology. On the other hand, AVIF uses AV1 compression technology, which is more advanced and efficient than WebP. Therefore, the ...
  27. [27]
    AVIF image format | Can I use... Support tables for HTML5, CSS3, etc
    Animated image sequences are not supported. 3 Only available on macOS 13 ... Usage share statistics by StatCounter GlobalStats for October, 2025.
  28. [28]
    Usage Statistics of AVIF for Websites, November 2025 - W3Techs
    AVIF is used by 1.0% of all the websites. Historical trend. This diagram shows the historical trend in the percentage of websites using AVIF. Our dedicated ...
  29. [29]
    The State of WebP Browser Support - KeyCDN
    Jan 6, 2023 · 96.3% of browsers support WebP, including Chrome, Firefox, and Edge. Only Internet Explorer 11 and KaiOS browser do not support it.
  30. [30]
    WebP Browser Support Guide - WebP to PNG Converter
    2012-2013: Early Adoption. Opera and Android browsers add WebP support. Chrome adoption grows. 2018: Microsoft Edge Support. Edge adds WebP support, covering ...
  31. [31]
    Downloading and Installing WebP - Google for Developers
    Aug 7, 2025 · Download precompiled WebP utilities and libraries for Linux, Windows, and macOS, including encoding/decoding tools, viewers, and muxing ...
  32. [32]
    Photoshop plug-in for opening and saving WebP images - GitHub
    Important note: Photoshop 23.2 and newer natively supports WebP. However some features such as preview at encoding and animations are missing in Photoshop 23.2.
  33. [33]
    Work with WebP files in Photoshop - Adobe Help Center
    Aug 13, 2024 · With Photoshop 23.2, Photoshop provides full support to the WebP file format. Open, create, edit, and save WebP files in Photoshop without a plug-in or ...Download Plug-In For... · Install Plug-In · Quarantine Plug-In (macos...
  34. [34]
    GIMP 2.10 Release Notes
    GIMP now also ships with native WebP support, including features like animation, ICC profiles, and metadata. Both importing and exporting are supported. The ...
  35. [35]
    WebP Filetype (2025-09-20) - Paint.NET Forum
    Sep 20, 2025 · This plugin allows Paint.NET to load and save Google's WebP format. This plugin is bundled with Paint.NET 4.2.5 and later.
  36. [36]
    paint.net 4.2.5 is now available!
    Oct 1, 2019 · paint.net 4.2.5 is now available! ... If you're using the Windows Store release, you should get the update automatically within the next 24 hours.
  37. [37]
    IrfanView PlugIns
    WEBP - (version 4.68): allows IrfanView to read/write WEBP (Weppy Format) files; WPG - (version 3.1.2): allows IrfanView to read WPG (version 1) files; WSQ ...Missing: XnView | Show results with:XnView
  38. [38]
    Webp support... - XnView Software
    Jan 28, 2021 · Webp not supported on classic.. updated the webp plugin and it still doesn't show webp files... meanwhile xnviewMP displays the same webp files just fine.
  39. [39]
    Affinity Photo 2 support of WebP format.
    Nov 13, 2022 · Hello, today I learned that "Affinity Photo 2" can import and export images in "Webp format". Unfortunately this format is not in the list.radical colour change when exporting webp - Affinity | ForumExport Persona, selecting an "Export setup preset" including Webp ...More results from forum.affinity.serif.comMissing: 2025 | Show results with:2025
  40. [40]
    Serif releases Affinity Photo 2.4 | CG Channel
    Feb 28, 2024 · Key changes in Affinity Photo 2.4 include the addition of 32-bit PNG files to the list of file formats in which HDR images can be imported and ...
  41. [41]
    What Is a WebP File? How WebP Compares To JPEG and PNG (2025)
    Mar 28, 2025 · WebP is a file format designed to optimize images for the web, improving loading times and enhancing overall website performance.
  42. [42]
    WebP Image Extension - Microsoft Apps
    Rating 3.0 (136) · Free · WindowsMay 23, 2018 · The WebP Image Extension will enable you to view WebP images in the Windows File Explorer, Photos and other Windows apps. WebP is a modern image ...
  43. [43]
    Safari 14 Release Notes | Apple Developer Documentation
    Sep 16, 2020 · New Features · Added WebP image support. · Added HDR video playback support. · Changed to derive <img> aspect ratio from size attributes. · Support ...
  44. [44]
    WebP Encoding Options - ImageMagick
    ImageMagick's WebP image format accepts a plethora of encoding options as detailed below. The default encoding options are:.
  45. [45]
    lovell/sharp: High performance Node.js image processing ... - GitHub
    This high speed Node-API module is to convert large images in common formats to smaller, web-friendly JPEG, PNG, WebP, GIF and AVIF images of varying ...
  46. [46]
    WordPress 5.8 adds WebP support
    Jun 7, 2021 · WebP images support lossy and lossless compression, as well as an animated format and support for transparent images. ... Date: May 5, 2021 . Adam ...
  47. [47]
    Telegram Stickers
    The built-in Sticker Editor lets everyday users seamlessly create stickers in the Telegram app with their own photos to share in chats. Artists and content ...Import API · Encoding Video Stickers and... · Interactive Custom Emoji
  48. [48]
    Modern Image Formats at Discord: Supporting WebP and AVIF
    Mar 12, 2025 · Recently Discord's Media Infrastructure team has modernized the app's image pipeline by adding support for animated WebP and AVIF formats!
  49. [49]
    Can I email… webP image format - CanIEmail
    webP support is 97.62% overall, but some clients have partial support, like converting to JPG or not supporting animation. Some webmail rendering depends on ...
  50. [50]
    Investigate email client support for WebP, plugin usage #74 - GitHub
    Dec 31, 2021 · Outlook has no support for WebP through its Email client on Windows, however the latest versions of Windows Mail, Outlook for Mac, Outlook.com, ...
  51. [51]
    Webp in gmail / transparency issues - Google Groups
    This is a known missing feature from GMail back end. It converts WebP images into JPEG ones for compatibility purposes.
  52. [52]
    [PDF] Accelerating the WebP Encoding/Decoding Pipeline - Emma Liu
    The project implemented a stage of the WebP encoding pipeline on NVIDIA GPUs, focusing on lossless compression and parallelizing the encoding stage.
  53. [53]
    WEBP Converter - CloudConvert
    CloudConvert converts your image files online. Amongst many others, we support PNG, JPG, GIF, WEBP and HEIC. You can use the options to control image resolution ...WEBP to JPG · WEBP to PNG · WEBP to GIF
  54. [54]
    Creating WebP Images with the Command Line | Articles - web.dev
    Nov 5, 2018 · The webp command line tool has already been installed for you, so you're all set to get started. This tool converts JPG, PNG, and TIFF images to WebP.Missing: non- double-
  55. [55]
    Why does Google, in it's Page Insights tool, advise to use WebP ...
    Apr 29, 2020 · Google has a vested interest in getting WebP supported by all major browsers so that they no longer have to serve up less efficient JPG and PNG images.
  56. [56]
    webp is actually crazy, why is widespread adoption so far behind?
    Oct 4, 2024 · A lot of CDNs also send webp content regardless of the file extension. Pop open a lot of those png and jpg files and you might find a webp.
  57. [57]
    WebP is so great… except it's not - Aurélien Pierre
    Oct 2, 2021 · First, at similar visual quality and for photographs, WebP is not lighter than JPEG, it's actually the opposite. All the Google claims rely on ...
  58. [58]
    The Battle to Dethrone JPEG with JPEG-XL, AVIF, and WEBP
    Feb 28, 2021 · In terms of decoding speed, JPEG-XL is the fastest with mozjpeg being very close, with WebP trailing behind alongside, and AVIF is last. With ...<|separator|>
  59. [59]
    WebP: A Personal Experiment | by Clio Labs
    Jun 14, 2013 · Encoding: 5–10x slower than JPEG (done once); Decoding: ~1.3x slower than JPEG (per decode). High encoding cost may be a limitation for use ...
  60. [60]
    Comparative Analysis on WebP and JPEG Decoding Performance
    The usual rule of thumb that we empirically found is that lossy WebP is 2/3 times slower to encode (with default settings) than JPEG, and 1.5 slower to decode.
  61. [61]
    AVIF vs WebP vs JPEG XL Comprehensive Comparison 2025
    Sep 20, 2025 · Photo content: AVIF achieves smallest size, 10–25% reduction at equivalent visual quality thresholds. Illustrations/screenshots: WebP ...
  62. [62]
    AVIF for Next-Generation Image Coding | by Netflix Technology Blog
    Feb 13, 2020 · Along with BD rates associated with various quality metrics, such as SSIM, MS-SSIM, VIF and PSNR, we also show rate-quality plots using SSIM as ...
  63. [63]
    (PDF) Web Image Formats: Assessment of Their Real-World-Usage ...
    Compared to the compressed JPEG format, WEBP and AVIF improved the Page Load Time by 21% and 15%, respectively. However, web scraping revealed that JPEG and PNG ...
  64. [64]
  65. [65]
    Uncovering the Hidden WebP vulnerability: a tale of a CVE with ...
    Oct 5, 2023 · Recently, on September 12, 2023, Google announced a security issue in Google Chrome, titled "Heap buffer overflow in WebP in Google Chrome," ...
  66. [66]
    CVE-2023-4863 and CVE-2023-5217 Exploited in the Wild | Wiz Blog
    Sep 11, 2023 · CVE-2023-4863 is a critical vulnerability in libwebp, and CVE-2023-5217 is a high severity vulnerability in libvpx, both reportedly exploited in the wild.
  67. [67]
  68. [68]
    CVE-2023-5129 - Red Hat Customer Portal
    Bugzilla 2240759: libwebp: out-of-bounds write with a specially crafted WebP lossless file ; CWE-122: Heap-based Buffer Overflow ; FAQ: Frequently asked questions ...
  69. [69]
    Critical libwebp Vulnerability - CVE-2023-5129 - eSentire
    The issue, tracked as CVE-2023-5129 (CVSS: 10), is a heap buffer overflow vulnerability in the open source libwebp code library.
  70. [70]
    CVE-2016-9085 : Multiple integer overflows in libwebp allows ...
    Feb 3, 2017 · This occurs when an integer value is incremented to a value that is too large to store in the associated representation. When this occurs, the ...Missing: root | Show results with:root
  71. [71]
    Blasting Past Webp - Google Project Zero
    Mar 26, 2025 · An analysis of the NSO BLASTPASS iMessage exploit. Posted by Ian Beer, Google Project Zero. On September 7, 2023 Apple issued an out-of-band ...
  72. [72]
    Security: Integer overflow in libwebp "ParseOptionalChunks" allows ...
    VULNERABILITY DETAILS An integer overflow in libwebp 'ParseOptionalChunks' (third_party/libwebp/dec/webp.c) allows an attacker to disclose the contents of ...Missing: root | Show results with:root
  73. [73]
    TALOS-2025-2224 || Cisco Talos Intelligence Group
    Aug 25, 2025 · When loading a specially crafted .webp animation an integer overflow can be made to occur when calculating the stride for decoding. Afterwards, ...Missing: root | Show results with:root
  74. [74]
    Resolving Webp Zero-day Vulnerability CVE-2023-4863 | Veracode
    Sep 29, 2023 · The webp image library is vulnerable to Heap Buffer Overflow. The exact steps to exploit the vulnerability have not been disclosed publicly.
  75. [75]
    Heap-based Buffer Overflow in electron | CVE-2023-4863 | Snyk
    Sep 12, 2023 · Affected versions of this package are vulnerable to Heap-based Buffer Overflow when the ReadHuffmanCodes() function is used.
  76. [76]
    Patching High Impact Vulnerabilities: A Retrospective on WebP CVE
    May 13, 2024 · The WebP vulnerability is exploited by crafting an image that causes the parser to overflow a buffer on the heap and distributing it to the ...Missing: gallery | Show results with:gallery<|separator|>
  77. [77]
    Advisory: libwebp critical vulnerability - Sophos
    Oct 3, 2023 · Overview. On Wednesday September 13, 2023, the WebP project released version 1.3.2 of libwebp containing a fix for a critical severity ...
  78. [78]
    Site Isolation - The Chromium Projects
    Site Isolation is a security feature in Chrome that offers additional protection against some types of security bugs.
  79. [79]
    Breach to Barrier: Strengthening Apps with the Sandbox - Electron
    Sep 29, 2023 · The sandbox puts a hard cage around all renderer processes, ensuring that no matter what happens inside, code is executed inside a restricted environment.
  80. [80]
    Picture this Exploit: Exploring the WebP Image Vulnerability CVE ...
    Sep 28, 2023 · This is called a heap buffer overflow, a sub-class of the very common buffer overflow vulnerability. Attackers can craft malicious WebP images ...Missing: 2011- 4653
  81. [81]
    Google Is Already Experimenting With WebP2 As Successor To ...
    Nov 15, 2020 · WebP 2 is the successor of the WebP image format, currently in development. It is not ready for general use, and the format is not finalized.
  82. [82]
    WebP 2
    - **Development**: WebP 2 is an experimental image codec based on WebP, used for compression experiments, not a released format.
  83. [83]
    codecs/libwebp2 - Git at Google
    WebP 2 is an experimental image codec based on WebP. WebP 2 will not be released as an image format but is used as a playground for image compression ...
  84. [84]
    WebP2 Image Encode Benchmark - OpenBenchmarking.org
    Jan 26, 2021 · WebP2 supports 10-bit HDR, more efficienct lossy compression, improved lossless compression, animation support, and full multi-threading support ...