Attributions & Licenses
Third-party model and library attributions for TabTasker, including upstream pages and license references.
TabTasker is built on top of dozens of permissively-licensed open-source projects and machine-learning models. This page lists the upstream libraries, models and assets we depend on, the license each one is distributed under, and a link to its source so you can review or reuse them yourself.
On-device AI features use models like OpenAI Whisper for speech-to-text, SpeechT5 for text-to-speech, CLIP for image tagging, BLIP for captioning and TensorFlow.js MediaPipe for face detection. They run locally through transformers.js, ONNX Runtime Web and the WebGPU backend so no audio, image or text leaves your browser when you use them.
PDF tooling is powered by @cantoo/pdf-lib (an actively-maintained pdf-lib fork) and a custom WebAssembly compressor. Image tooling uses jSquash codecs for AVIF, JPEG and WebP encoding, plus UpscalerJS for super-resolution. Audio editing builds on the Web Audio API and lamejs / ffmpeg.wasm for export.
If you spot a missing credit or an outdated license reference, please let us know through the contact page and we will update this list. Full per-package licenses are also bundled with the deployed build under /licenses/.