Machine-learning apps that run entirely in your browser — no server, no API keys.
Classifies text as positive or negative entirely in your browser.
Classify any text into custom labels with no training needed.
Identifies people, places, organisations and more in raw text.
Extractive QA that finds the answer inside a context passage.
Condenses long text into a short summary.
Translates text between English and 13 languages.
Predicts the missing word in a sentence.
Classifies an image into 1000 object categories.
Detects and localizes objects in an image.
Segments every object and region in an image.
Per-pixel depth from a single image.
Matches an image to free-text labels, no training needed.
Detects objects by free-text description, no fixed label set.
Generates a natural-language description of an image.
Answers questions about a document image (receipts, forms).
Transcribes speech fully in your browser.
Classifies sounds, music, and environments.
Synthesizes natural-sounding speech from text.
Embedding-based search over text, ranked by semantic similarity.
Retrieval-augmented generation pipeline running client-side.
A small chat LLM running locally in the browser.