Omebly Tools
Open-source credits
Our free tools run in your browser and are built on the work of many open-source projects. Thank you to everyone who made them. Licence files are linked from each project's page.
Libraries
- Astro · MIT
- Transformers.js (Hugging Face) · Apache-2.0
- ONNX Runtime Web (Microsoft) · MIT
- PDF.js (Mozilla) · Apache-2.0
- pdf-lib and fontkit · MIT
- qpdf-wasm · ISC
- Tesseract.js · Apache-2.0
- Mediabunny · MPL-2.0
- ag-psd · MIT
- libheif-js · LGPL-3.0
- libraw-wasm (LibRaw) · ISC; LibRaw under LGPL-2.1 or CDDL-1.0
- jSquash AVIF · Apache-2.0
- UTIF.js · MIT
- gifenc · MIT
- JSZip · MIT
- RNNoise (WebAssembly build by Shiguredo) · Apache-2.0; RNNoise BSD-3-Clause
- docx, docx-preview, PptxGenJS, pptx-preview, ExcelJS · MIT, Apache-2.0, MIT, ISC, MIT
- sanitize-html · MIT
- LiveKit client and server SDK (Omebly Meet) · Apache-2.0
- LiveKit track processors (Omebly Meet backgrounds) · Apache-2.0
- MediaPipe Tasks Vision (Google) · Apache-2.0
- Excalidraw (Omebly Meet whiteboard) · MIT
- React · MIT
- web-noise-suppressor (Omebly Meet noise removal) · MIT; RNNoise BSD-3-Clause
- qrcode-generator · MIT
- GSAP · GSAP Standard No-Charge License
- Lenis · MIT
AI models (run on your own device)
- ISNet (DIS), background removal · Apache-2.0 · Xuebin Qin and the DIS authors
- Real-ESRGAN, photo enlarging · BSD-3-Clause · Xintao Wang and the Real-ESRGAN authors
- GFPGAN v1.4, face restoring · Apache-2.0 · Tencent ARC Lab
- RetinaFace MobileNet-0.25, face finding · MIT · facexlib authors
- MI-GAN, object erasing · MIT · Picsart AI Research
- PP-OCRv6 small, text reading · Apache-2.0 · PaddlePaddle (Baidu)
- SlimSAM-77, click to select · Apache-2.0 · Meta AI (Segment Anything) and the SlimSAM authors
- Depth Anything V2 Small, depth for blur · Apache-2.0 · TikTok and the University of Hong Kong
- RTMDet-tiny, people and object finding · Apache-2.0 · OpenMMLab
- MediaPipe Face Mesh V2, face points · Apache-2.0 · Google
- MediaPipe Selfie Segmenter, Omebly Meet backgrounds · Apache-2.0 · Google
- LaMa (big-lama), large fills · Apache-2.0 · Samsung AI Center (advimman)
- DDColor tiny, colourising · Apache-2.0 · DDColor authors (piddnad)
- AdaIN style transfer · MIT; VGG-19 weights CC BY 4.0 (Simonyan and Zisserman) · naoto0804/pytorch-AdaIN
- Grounding DINO tiny, select by text · Apache-2.0 · IDEA Research
- BERT uncased WordPiece vocabulary · Apache-2.0 · Google
Fonts and data
- Open fonts from the Google Fonts collection (loaded from jsDelivr only when you turn on the free font list) · SIL Open Font License 1.1, Apache-2.0 and the Ubuntu Font Licence, as each font says
- English (US) spelling dictionary, from SCOWL and Hunspell (en_US 2020.12.07) · LGPL / MPL / BSD-style, see the dictionary licence file
- English (US) hyphenation patterns (the hyphen package, from TeX hyph-en-us by Gerard D.C. Kuiken) · ISC
- Lensfun lens database · CC BY-SA 3.0
- GL Transitions (video transitions) · MIT, as each transition says
- Manrope and JetBrains Mono (Omebly Meet), the Excalidraw whiteboard fonts · SIL Open Font License 1.1
Licence texts that come with these files: dictionary, hyphenation, Lensfun, LibRaw, GL Transitions.