![]()
Drag and drop WebP images. Transparency and animation detected automatically.
Choose page size, orientation, and whether to preserve transparency (lossless) or optimize for size (lossy).
Click Convert. Download your PDF. Files deleted immediately after download.
Modern image format by Google. Up to 30% smaller than JPG with the same quality.
Portable Document Format. Convert PDF pages to JPG, PNG or WebP images.
Apple photo format used by iPhone and iPad. High quality with small file size.
High Efficiency Image Format — same as HEIC, used on Apple devices.
Lossless format that preserves every pixel. Best for screenshots and logos.
Universal format for photos. Supported everywhere, great balance between quality and file size.
Classic format for simple animations. Supports transparency and up to 256 colors.
Uncompressed bitmap format. Maximum quality but very large file size.
Professional lossless format used in printing and photography.
Next-gen format with excellent compression. Up to 50% smaller than JPG.
Portable Pixmap format used in Unix/Linux environments.
High Dynamic Range format storing extended brightness data.
Flexible Image Transport System used in astronomy and science.
Quick comparison to help you choose the right format
Chrome and Firefox serve WebP via content negotiation — when you right-click and Save Image, you get a .webp file even if the website originally uploaded a JPEG or PNG. Google Images, Pinterest, Wikipedia thumbnails, Behance previews, and any site using WordPress with ShortPixel, Smush, or mod_pagespeed delivers WebP automatically. AI image generators made the problem bigger: DALL-E in ChatGPT switched from PNG to WebP downloads for C2PA content watermarking compliance. Midjourney images downloaded from Discord arrive as .webp. Stable Diffusion with AUTOMATIC1111 supports WebP as an output format. ComfyUI ships a native SaveAnimatedWEBP node. The result: a format that browsers handle perfectly but that the rest of the software ecosystem — Acrobat, Word, PowerPoint, email clients, government portals, print services — largely cannot.
WebP is an exceptionally efficient format. Lossy WebP is 25–34% smaller than JPEG at equal quality. Lossless WebP is up to 26% smaller than PNG. But the PDF specification has no WebPDecode filter — there is no way to store WebP data natively inside a PDF the way JPG is stored via DCTDecode. Every WebP must be decoded and re-encoded into a format PDF understands: DCTDecode (JPEG re-encode) for lossy content, or FlateDecode (PNG-style) for lossless content. This re-encoding typically inflates the file by 1.5× to 5× — a 200 KB WebP becomes a 300–500 KB page in the PDF. For a single image this is negligible. For 50 AI generations compiled into a portfolio PDF, the inflation is significant. The tradeoff is universal compatibility: every PDF reader on every platform can display the result.
Midjourney: generate variations on Discord, download the upscaled .webp file, upload to Convertify, combine 20–30 generations into a concept-art lookbook PDF for client review. DALL-E: ChatGPT now delivers WebP downloads — save them from the conversation, batch-convert to PDF for portfolio presentation or contest submission. Stable Diffusion (AUTOMATIC1111): set samples_format to webp for space-efficient storage of generation runs, then batch-convert selected outputs to PDF for print or archive. ComfyUI: use the SaveAnimatedWEBP node for motion outputs, then convert individual frames to multi-page PDFs for frame-by-frame client review. The common thread: AI tools generate in WebP for efficiency, but clients, contests, print shops, and archival systems need PDF.
The conversion pipeline has three stages. First, the WebP file is decoded using libwebp: VP8 lossy frames become raw RGB pixels, VP8L lossless frames become raw RGBA pixels with alpha channel preserved. Second, the converter chooses an embedding strategy: for lossy photographic WebP it re-encodes to JPEG via DCTDecode (smaller PDF, one generation loss), for lossless or transparent WebP it encodes via FlateDecode (larger PDF, pixel-perfect, alpha preserved via SMask). Third, the PDF structure is written: page objects, Image XObjects with the appropriate filter, and content streams positioning each image. For animated WebP files with multiple ANIM/ANMF chunks, each frame is extracted and placed on a separate PDF page — creating a multi-page flipbook.
WebP supports 8-bit alpha in both lossy (VP8 + alpha sidecar) and lossless (VP8L native RGBA) modes. When the converter uses the FlateDecode path and writes a proper SMask XObject, the transparency is preserved in the PDF — a transparent WebP logo will composite correctly over any background in InDesign, Illustrator, or professional print software. When the converter uses the DCTDecode path (JPEG re-encode), transparency is destroyed because JPEG has no alpha channel. If your WebP has transparency that you need to keep, make sure the converter uses lossless embedding. Convertify auto-detects alpha channels and routes transparent WebPs through the FlateDecode path.
| Approach | Quality | File size | Transparency | Effort |
|---|---|---|---|---|
| WebP → PDF (direct) | Minimal loss or lossless | 1.5–5× WebP size | Preserved if FlateDecode | One step |
| WebP → JPG → PDF | Two lossy generations | ~same as direct DCT | Lost | Two steps |
| WebP → PNG → PDF | Lossless decode + lossless embed | 3–8× WebP size | Preserved | Two steps |
| Share WebP directly | Original quality | Smallest | Preserved | Zero — but recipients may not open it |
You can, but you lose twice. The WebP → JPG step introduces one generation of JPEG compression artifacts. The JPG → PDF step with a proper converter adds zero additional loss (DCTDecode wraps the JPG byte-for-byte), but the damage from step one is already done. If you convert WebP → PDF directly, the converter does the JPG re-encode internally in a single optimized step — same quality loss but fewer manual steps and less risk of accidentally double-compressing. For lossless WebP (pixel art, AI art that needs exact pixel preservation), the WebP → JPG → PDF chain is destructive. WebP → PDF via FlateDecode preserves every pixel.
Photoshop will not open .webp: upgrade to Photoshop 23.2+ or install the WebPShop plugin from Google. This is a Photoshop limitation, not a WebP problem. DALL-E suddenly downloads .webp instead of .png: OpenAI adopted C2PA content watermarking which uses WebP as the delivery format — this is permanent, not a bug. Midjourney files from Discord are .webp: expected behavior — Discord serves uploaded images as WebP. PDF is 3× larger than the original WebP: the file-size paradox — WebP compression is more efficient than anything PDF supports natively. Lossless WebP became blurry in PDF: the converter used DCTDecode (JPEG re-encode) instead of FlateDecode — Convertify auto-selects FlateDecode for lossless inputs. Animated WebP only shows first frame: most converters flatten animations — Convertify decomposes each frame into a separate PDF page. Transparency disappeared: converter used DCTDecode path — Convertify routes transparent WebPs through FlateDecode with SMask. Chrome keeps saving as .webp: this is content negotiation — the server sends WebP to compatible browsers. Cannot avoid it without browser extensions.