Everything runs locally in your browser. No uploads, no servers, no accounts.
Powered by WebAssembly and on-device AI.
Most online converters are limited, insecure, and boring. They only handle conversions within the same media type and force you to upload your files to some random server.
Morph.it is different — it processes everything locally using WebAssembly and on-device AI, supports cross-medium conversions, and packs a full suite of creative tools without ever touching a server.
Need to turn an AVI into a PDF? Extract text from a scanned document? Generate speech from text? Edit a PDF? Go for it.
Morph.it is a fork of p2r3/convert. The original project is a brilliant universal file converter — Morph.it builds on that foundation with 7 new tools, a redesigned UI, on-device AI, and major performance upgrades.
The original project focused on format conversion. Morph.it adds 7 new tools on top — including video compression, AI background removal, a full video editor, neural text-to-speech & speech-to-text, AI summarization, OCR, and a PDF editor. See the full breakdown in the Tools table below.
- Full settings modal with per-tool configuration panels
- Dark/light theme toggle
- 11 accent colors — 8 presets + 3 custom slots with full color picker
- Recent files history on the home page
- Output tray — collect converted files instead of auto-downloading, or toggle auto-download
- In-app error log viewer with color-coded levels
- A* pathfinding for conversions — finds the optimal multi-step route between any two formats using a graph-based traversal system, so even formats with no direct converter can be reached through intermediate steps
- Parallel batch conversion — multiple files convert simultaneously using
navigator.hardwareConcurrency, with per-handler serialization to prevent shared-state conflicts (e.g. FFmpeg's shared virtual filesystem) - Per-file cancel/skip — skip individual files during a batch without stopping the rest
- Dynamic memory detection — probes the browser's max
ArrayBuffersize and warns before conversions that could crash the tab - 3-tier CDN fallback — jsDelivr → unpkg → local, so library downloads succeed even if a CDN is down
- Cache management — full settings UI showing total cache size and file count, with refresh, clear cache, and "clear all site data" reset
- Persistent storage — requests
navigator.storage.persist()so the browser won't evict cached models
- Completion chime — Web Audio API synthesized ding when conversions finish
- Background notifications — system notification when conversion completes in a background tab
- Conversion timer — shows elapsed time in success popups
- Copy to clipboard — one-click copy for text-based outputs (OCR, summarization, speech-to-text)
- Keyboard shortcuts — Enter to convert, Escape to close modals/return home
- Privacy mode — strips EXIF/GPS metadata, randomizes filenames, hides referrer headers
- First-visit cache prompt — asks whether to cache AI models locally before any downloads
Check out more information on the tools in the table below ↓
Change files between images, video, audio, documents, archives, fonts, 3D models, game assets, and more. Auto-detects input format with batch conversion and category queueing. Simple mode for everyday use, Advanced mode for power users. |
Re-encode videos with quality control and target file size constraints. Supports H.264, H.265, VP9 codecs with presets for Discord, Twitter/X, or custom targets. Output as MP4 or WebM. |
Background removal via on-device RMBG-1.4 (WebGPU/WASM) or remove.bg API with correction mode for text and fine details. AI image generation and editing via OpenRouter API. Plus rescaling with aspect ratio lock and metadata stripping. |
Trim, crop, merge, and manage audio with a 5-band parametric EQ. Extract, burn, or AI-generate subtitles with Whisper in 15 languages. Hardware-accelerated WebCodecs where available. |
28 neural voices (Kokoro 82M) across American and British accents with speed control and fullscreen read-aloud mode. Speech-to-text with 4 Whisper model sizes and word-level timestamps. |
Summarize PDFs, DOCX, text, or web pages with DistilBART/BART models. Smart chunking handles long documents automatically. Adjustable target length from 50–500 words. |
Extract text from images and scanned PDFs in 14 languages. Multi-page PDF support with live preview. Includes fullscreen read-aloud mode with Kokoro TTS integration. |
6 tools: select, text (20 fonts with auto-style matching), draw, highlight, erase, and image insertion. Per-page undo/redo, zoom, and live thumbnail sidebar. |
Universal converter — 200+ formats with searchable format picker and category filters
| Category | Examples |
|---|---|
| Image | PNG, JPEG, WebP, GIF, SVG, TIFF, BMP, ICO, HEIF, AVIF, JP2, JXL, QOI, VTF, Aseprite, and 50+ more |
| Video | MP4, AVI, MKV, WebM, MOV, FLV, and 100+ FFmpeg formats |
| Audio | MP3, WAV, OGG, FLAC, AAC, MIDI, MOD, XM, S3M, IT, QOA, and more |
| Document | PDF, DOCX, XLSX, PPTX, HTML, Markdown, EPUB, RTF, LaTeX, ODT, and 50+ via Pandoc |
| Data | JSON, XML, YAML, CSV, SQL, SQLite, NBT (Minecraft) |
| Archive | ZIP, 7Z, TAR, TAR.GZ, GZ, LZH |
| 3D Model | GLB and other formats via Three.js |
| Font | TTF, OTF, WOFF, WOFF2 |
| Game | Doom WAD, Beat Saber replays (BSOR), Scratch 3.0 (SB3), Portal 2 (SPPD), Half-Life 2 (VTF) |
| Other | Base64, hex, URL encoding, Python turtle graphics, PE executables |
Full image editor — layers, brushes, effects, selections, text, shapes, and AI generation
|
100% client-side — all processing runs in your browser using WebAssembly. Your files never leave your device unless you explicitly opt into remove.bg API, OpenRouter API (AI image generation), or CORS proxy. Privacy mode strips EXIF/GPS metadata, randomizes filenames, and hides referrer headers. No accounts, no tracking, no uploads. |
- Dark and light themes
- 8 preset accent colors + 3 custom color slots with full color picker
- Configurable defaults for every tool
- Auto-download toggle or collect files in the output tray
| Component | Technology |
|---|---|
| Build | TypeScript + Vite |
| Video/Audio | FFmpeg WASM |
| Images | ImageMagick WASM |
| Documents | Pandoc |
| OCR | Tesseract.js |
| Text-to-Speech | Kokoro TTS (82M params) |
| Speech-to-Text | Whisper via Transformers.js |
| Summarization | DistilBART / BART via Transformers.js |
| Background Removal | RMBG-1.4 via Transformers.js |
| pdfjs-dist + Fabric.js + pdf-lib | |
| 3D Models | Three.js |
| Archives | 7z-WASM, JSZip, pako |
| Music Trackers | libopenmpt |
- Go to morph.it (or convert.utoggl.in)
- Pick a tool from the home screen or drop files anywhere
- Configure your options and hit the action button
- Download your result — or keep working