64 Free AI Tools

Chat with AI for free, directly in your browser. This tool uses WebLLM to run open-source language models locally on your device via WebGPU hardware acceleration. No signup, no API keys, no server calls, and no data leaves your browser — ever. Choose from 14 models ranging from ultra-light 135M to powerful 8B-parameter models, configure a custom system prompt and temperature, and start chatting instantly. Your conversations are never stored, transmitted, or read by anyone.

Real-time AI vision detection powered by Google MediaPipe — the same technology used in Google Meet, YouTube, and Android. Detect faces, hands, body poses, and objects using your webcam or uploaded images. All processing runs locally in your browser via WebAssembly and WebGL. No signup, no server, no API calls. Your camera feed never leaves your device.

Convert robotic AI-generated text into natural, human-like writing with our AI Text Humanizer. This tool expands contractions, adds academic transitions, replaces words with sophisticated synonyms, and restructures sentences to make your content sound more authentic and help avoid AI detection. All processing happens locally in your browser for complete privacy.

Create well-structured, high-quality prompts for any AI model using a visual form that encodes the best practices from OpenAI, Anthropic, and Google prompt engineering guides. Set the role or persona (role prompting), define a clear task with step-by-step instructions (chain-of-thought), add context and few-shot examples, choose an output format (JSON, Markdown, code, table, numbered list), set the tone (professional, casual, technical, academic), and add explicit constraints to control scope and safety. The tool assembles everything into a copy-ready prompt optimized for ChatGPT (GPT-4o, GPT-4.5), Claude (Opus, Sonnet, Haiku), Gemini, Llama, Mistral, DeepSeek, and any other LLM or AI API endpoint. Supports zero-shot, few-shot, and chain-of-thought patterns. Runs entirely in your browser with no data ever leaving your device.

A complete retrieval-augmented generation (RAG) pipeline that runs entirely in your browser, with no backend and no vector database. It splits your document into overlapping chunks, embeds each one with the all-MiniLM-L6-v2 sentence-transformer (Xenova/all-MiniLM-L6-v2) via Hugging Face Transformers.js, holds the vectors in memory, and uses cosine similarity to retrieve the passages most relevant to your question. Those passages become the context for a local language model (Llama 3.2 1B) served by WebLLM, the open-source high-performance in-browser LLM inference engine from mlc-ai, which streams a grounded, source-cited answer. Because both the embedding model and the language model load and run on your own machine, through WebAssembly for embeddings and WebGPU for generation, your document never leaves your device. No API keys, no accounts, no server.

Upload or paste an image and instantly extract all text from it using Tesseract.js, the most popular open-source OCR engine. Supports 100+ languages including English, Arabic, Chinese, Japanese, Korean, Hindi, and more. All processing happens locally in your browser — no signup, no server, no API calls. Your images stay on your device.

Erase the background from any image and export a transparent PNG without sending a single pixel to a server. This tool runs the BRIA RMBG-1.4 image-segmentation model directly inside your browser with Hugging Face Transformers.js, which brings state-of-the-art machine learning to the web and runs the model on ONNX Runtime through WebAssembly with no need for a backend. Drag in a photo, load the model once (cached afterward), and download a cut-out subject with a clean alpha channel in seconds. No accounts, no credits, no watermarks, and complete privacy because nothing is ever uploaded.

Type or paste any text and instantly convert it to natural-sounding speech using Kokoro, an open-weight 82-million parameter AI voice model. Choose from 28 English voices across American and British accents with male and female options, and adjust speaking speed from 0.5x to 2x. Everything runs locally in your browser using WebAssembly — no signup, no server, no API calls. Your text never leaves your device.

A fast, private image format converter that runs entirely in your browser using jSquash, the open-source collection of WebAssembly image codecs derived from Google Squoosh. The same battle-tested codecs that power the Squoosh app, MozJPEG for JPEG, libwebp for WebP, and libavif for AVIF, run locally in your tab to encode, decode, and re-compress images with no native dependencies. Convert PNG and JPEG photos to modern formats like WebP and AVIF to dramatically shrink file size with no visible quality loss, or convert back to lossless, optimized PNG. A quality slider lets you trade size for fidelity on lossy targets, and a live size comparison shows exactly how many kilobytes you saved. Because the WASM codecs decode and encode on-device, your images never touch a server, never get logged, and never leave your machine.

Transcribe audio to text for free using OpenAI's Whisper speech recognition model running entirely in your browser. Record from your microphone or upload MP3, WAV, M4A, or other audio files and get accurate transcriptions in seconds. Supports 99 languages with automatic language detection, segment and word-level timestamps, and translation to English. No signup, no server, no API calls — all processing happens locally via WebAssembly using Hugging Face Transformers.js, so your audio never leaves your device.

A free, 100% client-side PDF extractor powered by pdf.js, the community-driven, web standards-based PDF parsing and rendering platform supported by Mozilla and built into Firefox since version 19. It reads your document directly in the browser, walking every page to pull out the embedded text layer and rendering each page to a crisp PNG image at 1.5x scale. Switch between a Text tab (concatenated text in reading order, with copy and download to .txt) and a Page images tab (per-page PNG thumbnails you can download individually) while a live progress indicator tracks each page. Because all parsing and rendering happen locally through the same Apache 2.0 licensed HTML5 engine that powers Firefox, even confidential contracts, invoices, and reports stay private on your machine. Scanned, image-only PDFs have no text layer, so the tool returns clean page images you can run through an OCR tool.

A free, private AI image captioner that describes any image and writes accessibility-ready alt text without sending your files anywhere. It runs the vit-gpt2-image-captioning vision-language model directly in your browser using Hugging Face Transformers.js, the library for state-of-the-art machine learning on the web, so your images never leave your device. Transformers.js runs the same pretrained models as the Python transformers library, executing them client-side through ONNX Runtime on WebAssembly (WASM) with optional WebGPU acceleration. Drag and drop a JPG, PNG, WebP, GIF, or BMP, load the model once (about 250MB, cached for next time), and get a natural-language caption you can copy as plain text or as a complete <img alt="..."> HTML snippet. Ideal for developers improving accessibility, content teams writing alt text at scale, and anyone who wants quick image-to-text descriptions without a cloud API or account.

A free, in-browser text to diagram editor powered by Mermaid, the open-source JavaScript diagramming and charting tool that turns Markdown-inspired text definitions into rendered diagrams. Write a few lines of simple, readable syntax and watch the diagram render instantly in the live preview. Start from prefilled templates for flowcharts, sequence diagrams, class diagrams, entity relationship (ER) diagrams, Gantt charts, state machines, and pie charts, then edit the code to match your system. The preview debounces as you type and surfaces friendly parser errors when the syntax is incomplete, so you never stare at a blank screen. When you are happy, copy the SVG markup to your clipboard or download a vector .svg for docs, slides, README files, and wikis. Everything runs locally in your browser with no server calls, so your architecture, schemas, and data never leave your device.

A free, private text summarizer that condenses long text into a concise summary without sending a single word to any server. It is built on Hugging Face Transformers.js, the open-source library that runs state-of-the-art machine learning directly in your browser with no need for a server. Under the hood it loads the Xenova/distilbart-cnn-6-6 model, a distilled BART fine-tuned on the CNN/DailyMail dataset, through the same pipeline() summarization API as the Python transformers library, executing it on your own hardware via ONNX Runtime and WebAssembly. Paste an article, meeting notes, a research paper, or a long email, drag a slider to control how long the abstract should be, and the model generates a fluent summary entirely on-device. Because every token is processed locally, your text is never uploaded, logged, or stored. The model downloads once on first use and is cached by your browser for instant summaries afterward.

A free, private paraphrasing tool that rewrites your text in different words while keeping the meaning, without sending a single word to any server. It is built on WebLLM (MLC AI), the open-source engine that runs full large language models directly in your browser using WebGPU for hardware acceleration. Under the hood it loads Meta Llama 3.2, a modern instruction-tuned model, and runs it entirely on your own GPU. Paste a sentence, a paragraph, or a whole essay, pick a rewriting style (standard, fluent, formal, simple, or concise), and the model produces a reworded version on-device. Because every token is generated locally, your text is never uploaded, logged, or stored. The model downloads once on first use and is cached by your browser, and it is the same model shared by the other in-browser AI tools, so it is ready instantly after the first download.

A privacy-first redaction tool that scrubs personally identifiable information from any text before it leaves your hands and reaches an AI chatbot. It works in two layers. The first is an instant regex pass that needs no model and zero download: it finds email addresses, phone numbers, credit-card numbers (validated with the Luhn checksum to cut false positives), IPv4 addresses, US Social-Security-style numbers, and long digit sequences such as account or order IDs. The second is an optional, on-demand layer that loads a small BERT named-entity-recognition model directly in your browser using Hugging Face Transformers.js (the library that lets you run Transformers directly in your browser, with no need for a server) to catch people, organizations, and locations that no regex can reliably find. Inference runs on the ONNX Runtime via WebAssembly (WASM), with optional WebGPU acceleration on supported devices. Every detected value is replaced with a stable, labeled token like [EMAIL_1] or [PERSON_2], and the same value always maps to the same token so the text stays coherent. You see the original and redacted text side by side, a count of what was removed by type, and a full token map you can keep private to map the AI answer back to real values. Everything runs locally in your browser. Your text is never uploaded to any server.

A free, private grammar checker that corrects grammar, spelling, and punctuation without sending a single word to any server. It is built on WebLLM (MLC AI), the open-source engine that runs full large language models directly in your browser using WebGPU for hardware acceleration. Under the hood it loads Meta Llama 3.2, a modern instruction-tuned model, and runs it entirely on your own GPU. Paste an email, an essay, a message, or any text, choose to either correct mistakes only or correct and lightly improve readability, and the model returns the cleaned-up version on-device. Because every token is generated locally, your text is never uploaded, logged, or stored. The model downloads once on first use and is cached by your browser, and it is the same model shared by the other in-browser AI tools, so it is ready instantly after the first download.

A free, privacy-first machine translation tool that runs Meta AI NLLB-200 (No Language Left Behind) directly in your browser through Hugging Face Transformers.js, the open-source library for running state-of-the-art machine learning on the web with no server. Pick a source and target language from 15 common options including English, Spanish, French, German, Arabic, Chinese, Hindi, Portuguese, Russian, Japanese, Korean, Italian, Dutch, Turkish, and Polish, then translate any text without sending it anywhere. Transformers.js loads the NLLB ONNX model and runs it on-device via WebAssembly, so the only network request is the one-time model download (around 600MB) which the browser then caches for instant reuse. Because everything happens locally, it works offline after the initial load and is well suited to confidential documents, regulated industries, and air-gapped or restricted networks where pasting text into a cloud service is not an option.

A free, private QR code and barcode scanner that runs entirely in your browser. It is built on html5-qrcode, a lightweight and cross-platform QR Code and Bar code scanning library for the web, so the same decoding engine used inside thousands of production apps powers this page. Use your phone or laptop camera for live scanning, or drop in a saved screenshot or photo to decode it instantly. The tool recognizes QR codes plus a wide range of 1D and 2D barcode formats including AZTEC, CODE_39, CODE_93, CODE_128, ITF, EAN_13, EAN_8, PDF_417, UPC_A, UPC_E, DATA_MATRIX, MAXICODE, RSS_14, and RSS_EXPANDED. It automatically detects when a result is a link and renders it as a clickable URL, and it keeps a running history of everything you scan so you can copy each value with one click. No images are uploaded, no accounts are required, and nothing is logged. The camera feed and decoding both happen on your device.

A free, private subtitle and caption generator that turns audio or video into timestamped SRT and WebVTT files without uploading anything. It is built on Hugging Face Transformers.js, the open-source library that runs state-of-the-art machine learning directly in your browser, and it loads OpenAI Whisper, the most widely used open-source speech recognition model. Drop in an MP3, WAV, M4A, MP4, WebM, or MOV file and the tool decodes the audio track, runs Whisper locally with segment-level timestamps, and produces ready-to-use .srt or .vtt subtitles you can preview, copy, or download. Because the model runs on your own device via WebAssembly, your media is never uploaded, logged, or stored, which makes it safe for unpublished videos and confidential recordings. The model downloads once on first use and is cached by your browser, and it shares that cache with the speech-to-text tool.

A free, private sentiment analyzer that scores the tone of text without sending a single word to any server. It is powered by Hugging Face Transformers.js, the JavaScript library that runs the same pretrained models as the Python transformers library, here executing the DistilBERT SST-2 sentiment model directly in your browser through ONNX Runtime and WebAssembly. Paste a product review, a support ticket, a survey response, or a batch of comments, and the tool returns positive or negative with a calibrated confidence score, rendered as clear, labeled horizontal bars. Switch to the optional RoBERTa GoEmotions model to detect 28 fine-grained feelings like joy, anger, sadness, fear, and gratitude. Turn on per-line mode to score many rows at once, then copy or download the full report. Because everything runs locally on your own hardware, your text is never uploaded, logged, or stored. Each model downloads once on first use and is cached by your browser for instant repeat analysis, and once cached the tool keeps working offline.

A free, private voice notes tool that turns spoken audio into a written transcript and a short AI summary without sending a single byte to any server. It chains two open-source models that run entirely on-device through Hugging Face Transformers.js: OpenAI Whisper for speech recognition and DistilBART CNN for abstractive summarization. Record straight from your microphone or upload an existing audio file, and the tool first transcribes the speech with the whisper-base model, then condenses that transcript into a concise summary with the Xenova/distilbart-cnn-6-6 model, using the same pipeline() API as the Python transformers library. Both models execute on your own hardware via ONNX Runtime and WebAssembly, so your audio and the resulting text are never uploaded, logged, or stored. The models download once on first use and are cached by your browser, and they are the same files shared with the speech-to-text and text-summarizer tools, so if you have used those, voice notes start almost instantly.

A free, private JSON to TypeScript converter that runs entirely in your browser. Paste any JSON object or array and instantly generate strongly typed TypeScript interfaces, runtime-validating Zod schemas, or a standards-compliant JSON Schema. It is powered by quicktype, the open-source engine that generates strongly typed models and serializers from JSON, JSON Schema, TypeScript, and GraphQL queries, and the same quicktype-core library used in editors, CLIs, and production build pipelines runs right here on this page. quicktype infers the type of every field from your sample data, including nested objects, arrays, optional and nullable properties, unions, and enums, then renders clean, idiomatic output you can paste straight into your codebase. Choose a root type name, switch between TypeScript, TypeScript plus Zod, and JSON Schema targets, validate your JSON with friendly inline errors, and copy or download the result in one click. No JSON is uploaded, no account is required, and nothing is logged: the parsing and code generation both happen on your device.

A free, private voice translator that records your microphone, transcribes your speech, translates it into another language, and can read the translation back to you, all without sending a single byte to any server. It chains three open-source models inside your browser tab: OpenAI Whisper handles speech recognition, Meta NLLB-200 handles translation across 15 common languages, and Kokoro handles text to speech for the result. Speak into the mic or upload an audio file and Whisper turns it into text on your own hardware; or skip audio entirely and just type the source text. NLLB-200 then translates that text into your target language, and an optional Kokoro pass speaks the translation aloud and lets you download it as a WAV file. Everything runs through Hugging Face Transformers.js and kokoro-js, executed on your device via ONNX Runtime and WebAssembly, so your voice and text are never uploaded, logged, or stored. The models download once on first use and are cached by your browser, and they are shared with the standalone speech-to-text, translator, and text-to-speech tools so you never download them twice.

A free, private CSV to JSON and JSON to CSV converter that runs entirely in your browser. It is built on Papa Parse, the fastest in-browser CSV (or delimited text) parser for JavaScript, which is reliable and correct according to RFC 4180 and has no dependencies. Paste your data or upload a .csv or .json file, then flip a single toggle to convert in either direction. When going from CSV to JSON, the tool uses the header row to turn each line into a keyed object and can auto-detect numbers and booleans so 42 becomes a number and true becomes a boolean instead of plain strings. When going from JSON to CSV, it flattens an array of objects back into clean, RFC 4180 compliant CSV with properly quoted fields and escaped delimiters. You get a live table preview of the first rows, the full converted output in a text box, clear error messages whenever the input cannot be parsed, and one-click copy to clipboard plus download of the result. No data is uploaded, no account is required, and nothing is logged. Parsing and conversion both happen on your device.

A free, private voice typing tool that transcribes your speech into an editable notepad in real time, without uploading a single second of audio. It is built on Hugging Face Transformers.js, the open-source library that runs machine learning models directly in your browser, and it loads OpenAI Whisper, the most widely used open-source speech recognition model. Click Start, grant microphone access, and the tool captures your voice in short rolling segments, runs Whisper locally every few seconds, and writes the growing transcript into a text area you can correct as you go. Because the model runs on your own device through WebAssembly, your voice is never uploaded, logged, or stored, which makes it safe for private notes, medical or legal dictation, and confidential drafting. The model downloads once on first use and is cached by your browser, and it shares that cache with the speech-to-text and subtitle-generator tools.

A free, private audio converter that changes audio files between MP3 and WAV entirely on your own device. It uses the browser Web Audio API to decode your file, which handles common formats like MP3, WAV, M4A, OGG, FLAC, and AAC, then re-encodes the raw audio to the format you pick. WAV output is written natively as standard 16-bit PCM, and MP3 output uses lamejs, the LAME MP3 encoder ported to JavaScript, so nothing is ever uploaded. Choose your output format, pick an MP3 bitrate from 128 to 320 kbps, and download the converted file with an in-page preview player. Because everything runs client-side, there are no file-size caps, no queues, no watermarks, and no ads sitting between you and your audio, unlike the typical online converter that uploads your file to a server first.

A free, private QR code generator that runs entirely in your browser. It is built on node-qrcode (soldair/node-qrcode), a QR code and 2D barcode generator that works on the server, the client, and React Native, so the same encoding engine used in thousands of production apps powers this page. Pick a content type and the tool builds the correct payload for you: a plain text or URL code, a WiFi join code (WIFI:T:WPA;S:ssid;P:password;;) that connects a phone to your network in one scan, a mailto: email code, an smsto: text-message code, or a full vCard contact card. Tune the output with a size slider, choose an error-correction level (L, M, Q, or H), and set custom foreground and background colors. A live preview re-renders as you type, and you can download the result as a crisp high-resolution PNG or as an infinitely scalable SVG that stays sharp at any print size. node-qrcode supports Numeric, Alphanumeric, Kanji and Byte modes plus Chinese, Cyrillic, Greek, Japanese and multibyte characters like emojis, and it auto-generates optimized segments for the best data compression. Nothing you enter is ever uploaded: the entire QR code is generated locally on your device, with no account, no API key, and no tracking.

A free, private audio trimmer and cutter that lets you slice a section out of any audio file without uploading a single byte. It is built on the browser Web Audio API, which decodes your file and gives direct access to the raw audio samples, so trimming is just a matter of copying the samples between your chosen start and end points. Drop in an MP3, WAV, M4A, OGG, or FLAC file, drag two sliders or type exact timestamps down to the millisecond, preview the original, then export the selected slice as an MP3 (with a bitrate of your choice) or a lossless WAV. MP3 encoding is handled by lamejs, a pure-JavaScript MP3 encoder that runs entirely in your browser, and WAV is written natively from the raw PCM samples. Because every step happens on your own device, there is no upload, no server processing, no watermark, no length limit, and no file-size cap. It is a fast, no-nonsense replacement for online audio cutters that make you wait for uploads and then stamp their brand on your file.

A free, private PDF merge and split tool that runs entirely in your browser. It is built on pdf-lib, an open-source library to create and modify PDF documents in any JavaScript environment, so the same engine trusted in thousands of production apps powers this page. In Merge mode you can add several PDF files, drag them into any order, see each file page count, and combine them into a single document. In Split mode you load one PDF and extract a page range like "1-3,5,8-10" into a brand new PDF, with full page reordering supported because you control the exact order pages are listed. Unlike pdf-lib peers that only generate PDFs, pdf-lib can both create and modify existing documents, which is exactly what merging, splitting, and reordering require. It is pure JavaScript with no native dependencies, so nothing is uploaded: your files are read with the browser FileReader, processed in memory, and the result is handed back as a downloadable application/pdf Blob. No accounts, no watermarks, no page limits, and nothing is logged or transmitted. Everything happens on your device, and you can prove it by watching the Network tab in DevTools while you work.

A free, private image compressor and resizer that runs entirely in your browser using browser-image-compression, an open-source JavaScript module made to run in the web browser for image compression. The same library used in thousands of production apps to shrink photos before upload powers this page, so compression happens on-device with no server round trip. Drag and drop one image or a whole batch of JPEG, PNG, or WebP files, then set a target maximum file size in megabytes and an optional maximum width or height to downscale oversized images. The library compresses by iteratively reducing quality and, when needed, resolution until your size target is met, and it runs that work in a Web Worker so the page stays responsive even with large photos. For every file you see the original size, the compressed size, and the exact percent saved, plus a side-by-side preview. Download each result individually or grab them all at once. No images are uploaded, no accounts are required, and nothing is logged: every byte stays in your browser tab.

A free, private text to audiobook generator that turns long text into a single downloadable MP3 narration without sending a single word to any server. It is powered by Kokoro, an open-weight 82 million parameter text-to-speech model, run entirely in your browser through the kokoro-js library and WebAssembly. Paste an article, a book chapter, meeting notes, or study material, pick one of nearly thirty natural English voices, and the tool splits your text into sentence-sized chunks, narrates each one on your own device, concatenates the audio, and encodes it into one clean MP3 you can preview and download. Because every word is processed locally, your text is never uploaded, logged, or stored. The Kokoro model is roughly 300MB, downloads once on first use, is cached by your browser for instant reuse, and shares that cache with the text-to-speech tool on this site. Kokoro is released under the permissive Apache 2.0 license, and its voices are English-style, spanning American and British accents in both female and male options.

A free, private audio noise reducer that removes background noise from voice recordings without uploading anything. It is built on RNNoise, the open-source noise-suppression model from Xiph.Org (the makers of Opus and Vorbis), compiled to WebAssembly so it runs entirely in your browser. RNNoise combines classic digital signal processing with a small recurrent neural network trained to tell speech from noise, and it works on 48kHz audio in real time. Upload an MP3, WAV, or M4A recording and the tool decodes it, resamples it to 48kHz mono, runs every frame through RNNoise, and gives you a cleaned file to preview and download as WAV or MP3. Because the model is tiny and runs on your own device, your recording is never uploaded, logged, or stored, which makes it safe for private interviews, calls, and voice memos. Nothing leaves the browser.

A free, private EXIF metadata viewer and remover that runs entirely in your browser. Reading is powered by exifr, described by its author Mike Kovarik as the fastest and most versatile JavaScript EXIF reading library, so this page extracts metadata from JPEG, TIFF, HEIC, AVIF, and PNG images in about a millisecond by reading only the bytes it needs. It surfaces every tag it can find: EXIF camera settings, GPS latitude and longitude, capture timestamps, camera make, model and serial number, lens information, editing software, and XMP and IPTC fields such as artist, copyright, and captions. The GPS coordinate is highlighted as a privacy risk with a one-click link to an exact map location, and device and identity fingerprints are called out separately. Because exifr reads metadata but cannot write or delete it, this tool removes metadata a different way: it re-encodes your image through an HTML canvas, drawing the decoded pixels into a fresh JPEG or PNG that carries no EXIF, GPS, XMP, or IPTC segments at all. You get a clean copy to download. Nothing is uploaded, no account is required, and nothing is logged: both the parse and the re-encode happen on your device.

A free, private AI image upscaler that enlarges and sharpens photos 2x without uploading anything. It is built on Hugging Face Transformers.js, the open-source library that runs state-of-the-art machine learning directly in your browser, and it loads Swin2SR, a Swin Transformer super-resolution model trained to reconstruct fine detail rather than just stretch pixels. Drop in a JPG, PNG, or WebP image and the tool runs Swin2SR locally to produce a version at twice the width and height, recovering edges and texture that a plain resize would leave blurry. Because the model runs on your own device via ONNX Runtime and WebAssembly, your images are never uploaded, logged, or stored, which makes it safe for private photos, unpublished artwork, and confidential screenshots. The model downloads once on first use and is cached by your browser, so repeat upscales are fast. To keep in-browser memory reasonable, input images are capped at 1024 pixels on the longest side.

A free color converter and WCAG contrast checker that runs entirely in your browser with zero dependencies. Convert any color between HEX, RGB, and HSL using the exact standard color math, pick colors visually with native color pickers, and instantly see whether a foreground and background pair passes WCAG 2.x accessibility contrast requirements. The contrast ratio is computed the way the Web Content Accessibility Guidelines define it: each sRGB channel is linearized (c <= 0.03928 ? c / 12.92 : ((c + 0.055) / 1.055) ^ 2.4), relative luminance is derived as L = 0.2126R + 0.7152G + 0.0722B, and the ratio is (L1 + 0.05) / (L2 + 0.05) where L1 is the lighter color. You get a live swatch preview, the precise ratio, and clear pass or fail badges for AA (4.5:1 normal text, 3:1 large text) and AAA (7:1 normal text, 4.5:1 large text). Everything happens locally on your device: nothing is uploaded, nothing is logged, no account is needed, and it works fully offline once the page has loaded.

A free, private photo anonymizer that automatically finds faces in a picture and blurs or pixelates them so identities are protected before you share the image. It is built on Google MediaPipe and its BlazeFace short-range model, an open-source face detector that runs directly in your browser through WebAssembly and WebGL. Upload a JPG, PNG, or WebP and the tool draws it to a canvas, detects every face on your own device, and redacts each one with an adjustable blur radius or pixel size. Because detection and redaction both happen locally, the image is never uploaded, logged, or stored, which makes it safe for personal photos, screenshots, client work, and street photography. You can switch between blur and pixelate, tune the strength with a slider without re-running detection, see how many faces were found, and download the anonymized result as a PNG. The MediaPipe model downloads once on first use and is cached by your browser afterward.

A free, private Unix timestamp converter that runs entirely in your browser with zero dependencies. It turns an epoch timestamp into a readable date and turns any date back into a Unix timestamp, both ways, instantly. Paste a value in seconds or milliseconds and the tool auto-detects the unit by digit count, then formats the result with the native JavaScript Intl.DateTimeFormat API in the time zone you pick. The time zone list is built from Intl.supportedValuesOf("timeZone") when your browser exposes it, merged with a small curated fallback list (UTC, your local zone, and common IANA zones) so it still works everywhere. A live clock shows the current Unix time in seconds and milliseconds, ticking once per second. All math is done internally in UTC milliseconds to avoid daylight-saving and locale drift, and every output is one click from your clipboard. Nothing is uploaded, no account is needed, and no value is logged. Because the conversion uses only the platform Date and Intl objects that ship with your browser, it is fast, accurate, and fully offline once the page has loaded.

A free, private depth map generator that turns any photo into a grayscale depth map without uploading anything. It is built on Hugging Face Transformers.js, the open-source library that runs state-of-the-art machine learning directly in your browser, and it loads Depth Anything, a leading monocular depth-estimation model that infers how far each pixel is from the camera from a single image. Drop in a JPG, PNG, or WebP and the tool runs the model locally, then renders a depth map where bright pixels are near the camera and dark pixels are far away. Because the model runs on your own device via ONNX Runtime and WebAssembly, your image is never uploaded, logged, or stored, which makes it safe for unpublished photos and confidential work. The model downloads once on first use (about 50MB) and is then cached by your browser, so later runs start instantly. You can view the original and the depth map side by side and download the depth map as a PNG for use in 3D, parallax, portrait and background effects, AI art with ControlNet, and photography workflows.

A free, private unique identifier generator that runs entirely in your browser with zero dependencies. It produces three of the most widely used ID formats: RFC 4122 version 4 UUIDs, lexicographically sortable ULIDs, and compact URL-safe NanoIDs. Every value is generated from cryptographically secure randomness using the native Web Crypto API (crypto.randomUUID and crypto.getRandomValues), the same CSPRNG browsers use for keys and tokens, so the IDs are suitable for production database primary keys, distributed systems, API request identifiers, and idempotency keys. Generate one ID or bulk-generate thousands at once, toggle UUID casing and hyphenation, choose a custom NanoID length, and rely on monotonic ULID generation that guarantees strictly increasing values even within the same millisecond. Copy any single ID, copy the whole batch, or download the results as a .txt file. No signup, no API keys, no network requests: nothing is uploaded, logged, or stored on a server. The generation happens locally on your device and works fully offline once the page has loaded.

A free, private URL encoder, decoder, and query string builder that runs entirely in your browser with zero dependencies. It uses the same standard JavaScript URL APIs your browser already ships, encodeURIComponent and decodeURIComponent for safe percent-encoding, encodeURI and decodeURI for whole-URL escaping that preserves reserved characters, and URL plus URLSearchParams for parsing and rebuilding query strings. Encode mode turns any text into a URL-safe string so spaces, ampersands, question marks, slashes, and unicode are escaped correctly. Decode mode reverses it, with a clear error when a string contains a malformed escape such as a dangling percent sign. A dedicated toggle exposes the classic space ambiguity: plus signs (the application/x-www-form-urlencoded form style) versus %20 (the RFC 3986 style), so you can match exactly what a server or framework expects. Query Builder mode parses an existing URL or bare query string into an editable key and value table, lets you add, edit, reorder by editing, and remove parameters, and rebuilds a correctly encoded URL live as you type. Everything is computed locally with native browser APIs: nothing is uploaded, no account is required, and nothing is logged.

A free, private AI cartoonizer that turns photos into anime and cartoon art without uploading anything. It is built on AnimeGANv2, the popular open-source photo-to-anime model, exported to ONNX and run entirely in your browser with ONNX Runtime Web on WebAssembly. Upload a photo, pick a style (Hayao for a soft Ghibli-like look, Shinkai for vivid color, or Paprika for bold stylization), and the model repaints your image in that style on your own device. Because everything runs locally, your photo is never uploaded, logged, or stored. Each style model is only about 9MB, downloads once, is cached by your browser, and runs on any modern browser without needing a GPU.

A free, private number base converter that runs entirely in your browser with zero dependencies. Type a value in any field - binary (base 2), octal (base 8), decimal (base 10), or hexadecimal (base 16) - and every other field updates live as you type, so you instantly see the same number expressed across all bases at once. There is also an arbitrary radix field where you can pick any base from 2 to 36 (digits 0-9 then A-Z) for both input and output. Conversion uses the browser native numeric engine: parseInt(str, radix) to read your input and Number.prototype.toString(radix) to render each base, the exact same primitives the JavaScript language ships. For whole numbers it uses BigInt internally so arbitrarily large integers convert exactly with no rounding or precision loss, far beyond the 53-bit safe-integer limit of a normal Number. Each base is validated against its own digit set, so an invalid digit (for example the letter G in hex, or a 2 in binary) shows a clear, friendly error instead of a silent wrong answer. Nothing is uploaded, no account is needed, and the page works fully offline once loaded. Copy any result to your clipboard with one click.

A free, private image cropper that runs entirely in your browser. It is built on Cropper.js (fengyuanchen/cropperjs), a JavaScript image cropper that, in the projects own words, supports 39 options, 27 methods, and 6 events, with touch and mobile support, so the same battle-tested engine used across thousands of production sites powers this page. Drag and drop a photo or click to browse, then drag the crop box to frame exactly the region you want. Lock the selection to a preset aspect ratio (Free, 1:1, 16:9, 4:3, 3:2, or 2:3) for a perfect square avatar, a widescreen banner, or a portrait thumbnail. Cropper.js exposes the full set of canvas transforms and this tool surfaces them: rotate the image in 90-degree steps, flip it horizontally or vertically, zoom in and out, and reset back to the original at any time. A live readout shows the exact pixel dimensions of the current crop as you adjust it. When you are happy, export through the librarys getCroppedCanvas method as a lossless PNG, or as a JPEG or WebP with an adjustable quality slider, then download the file or copy it straight to your clipboard to paste into another app. Cropper.js handles Exif orientation automatically, so photos taken on a phone are not rotated incorrectly. Nothing you load is ever uploaded: the image is read locally as an object URL and every pixel stays on your device, with no account, no API key, and no tracking.

A free, private LaTeX equation editor that renders TeX math entirely in your browser. It is powered by KaTeX (KaTeX/KaTeX), in the projects own words "a fast, easy-to-use JavaScript library for TeX math rendering on the web" whose layout is "based on Donald Knuth's TeX, the gold standard for math typesetting." Type LaTeX into the source box and a debounced live preview renders the equation synchronously, the same approach KaTeX uses in production on Khan Academy, GitHub, Discord, and thousands of other sites. Toggle between display mode (a centered block equation) and inline mode (text-flow size) to match how the formula will appear in your document. A grouped symbol palette lets you insert common structures with one click: fractions \frac{a}{b}, square and nth roots, superscripts and subscripts, big operators like \sum, \prod, \int and \lim, the full Greek alphabet, relations and arrows, and ready-made matrix, cases, and aligned environments. Five sample equations (the quadratic formula, the Gaussian integral, Euler's identity, a matrix equation, and the Basel sum) load instantly so you can see real output and learn the syntax. KaTeX renders math synchronously without reflowing the page and produces the same output across Chrome, Safari, Firefox, Opera, and Edge, so the preview you see is the preview everyone sees. When a parse error occurs, throwOnError surfaces the exact KaTeX error message in red while a best-effort preview still shows the rest of the expression. Export your work three ways: copy the raw LaTeX source, copy the rendered HTML span markup that KaTeX produces, or copy clean MathML for accessible documents and tools that consume it. KaTeX has no dependencies and runs locally, so nothing you type is ever uploaded: there is no account, no API key, and no tracking.

A free, private code formatter that runs entirely in your browser. It is powered by Prettier (prettier/prettier), the opinionated code formatter that enforces a consistent style by parsing your code and re-printing it with its own rules that take the maximum line length into account, wrapping code when necessary. The same engine that teams run on-save in their editor, in a pre-commit hook, or in CI to ensure a consistent codebase now runs locally on this page, with no install and no account. Pick a language and the tool selects the right Prettier parser and plugins for you: JavaScript and JSX (babel), TypeScript and TSX (typescript), JSON, CSS, SCSS, and LESS (postcss), HTML, Markdown, and YAML. Tune the output the way Prettier exposes it: set the print width (max line length), the tab width, whether to add semicolons, and whether to prefer single quotes. Click Format and the tool re-prints your code in milliseconds; if your input has a syntax error, Prettier reports exactly where, shown in a friendly error panel instead of failing silently. Copy the formatted result to your clipboard or download it with the correct file extension for the language you chose. Because Prettier ships a standalone browser build, every byte of your code is parsed and re-printed on your own device: nothing is uploaded, logged, or sent to any server. Prettier is open source and released under the MIT License.

A free, fully private two-way converter between HTML and Markdown that runs entirely in your browser. For HTML to Markdown it uses Turndown (mixmark-io/turndown), an open-source library whose own tagline is "Convert HTML into Markdown with JavaScript", trusted across thousands of production apps. For the reverse direction, Markdown to HTML, it uses marked, a fast, low-level Markdown compiler built for speed. Paste or type your source on the left, flip the direction toggle, and the converted output appears on the right alongside a rendered preview so you can see exactly what the result looks like. The HTML to Markdown engine is configurable the way Turndown is: ATX-style headings (# Heading), fenced code blocks with triple backticks, and a dash bullet-list marker by default, with Turndown automatically escaping Markdown special characters so your text never turns into unintended syntax. The Markdown to HTML engine uses marked with GitHub-flavored defaults so tables, fenced code, task lists, and inline formatting all render correctly. When you are happy with the result, copy it to the clipboard with one click or download it as a .md or .html file. Nothing you paste is ever uploaded: both Turndown and marked run as JavaScript on your device, with no account, no API key, and no server round-trip, which you can confirm in your browser DevTools Network tab.

Paste a JSON Web Token and instantly decode it to inspect the header, payload, and signature as defined by RFC 7519. See the algorithm (HS256, RS256, ES256), registered claims (sub, iat, exp, iss, aud, nbf, jti), expiration status, and issued date — all processed locally in your browser. JWTs are the backbone of modern authentication in OAuth 2.0, OpenID Connect, and API authorization flows, and this tool lets you debug them without exposing sensitive token data to third-party services. No server, no signup, no data sent anywhere — your tokens stay on your device.

Convert text to Base64 encoding or decode Base64 strings back to readable text, fully compliant with RFC 4648 which defines the standard Base64 alphabet of 64 printable ASCII characters plus = padding. Commonly used by developers for Basic Authentication headers, data URIs, email MIME encoding, JWT payloads, and embedding binary data in JSON or XML. Supports full UTF-8 including emojis and international characters. All processing runs locally in your browser — no server, no signup, no data sent anywhere — making it safe for encoding sensitive values like API keys, tokens, and credentials.

Paste YAML and get JSON, or paste JSON and get YAML. Powered by js-yaml, the most popular JavaScript YAML parser with tens of thousands of dependent packages on npm. Implements the YAML 1.2 specification — the version that fixed the infamous Norway problem and made YAML a strict superset of JSON. Handles nested objects, arrays, multiline strings (literal and folded block scalars), anchors, aliases, and all standard YAML features. Ideal for converting Kubernetes manifests, Docker Compose files, GitHub Actions workflows, Ansible playbooks, Helm charts, and any other YAML or JSON configuration. Runs locally in your browser — no data sent anywhere.

Format messy SQL into clean, readable queries using sql-formatter, the most widely used open-source SQL formatting library with over 750,000 weekly npm downloads. Supports 9 SQL dialects including MySQL, PostgreSQL, SQLite, Google BigQuery, SQL Server (T-SQL), Oracle (PL/SQL), Amazon Redshift, Apache Spark SQL, and standard SQL. Applies consistent formatting rules — uppercase keywords, proper indentation, aligned clauses, and logical line breaks — following best practices from style guides like Simon Holywell's SQL Style Guide. Handles complex queries with CTEs, subqueries, window functions, JOINs, and procedural blocks. Runs entirely in your browser via the sql-formatter TypeScript library, so your queries never leave your device. A free, private alternative to the SQL formatters built into paid tools like DataGrip, DBeaver, and Azure Data Studio.

Write a regex pattern and instantly see every match highlighted in your test string as you type. The tool displays full match details including numbered and named capture groups, match indices, and the effect of each flag. Supports all modern ECMAScript regex features — lookahead, lookbehind, named groups, Unicode property escapes, and the dotall flag. Six flag controls let you toggle global, case-insensitive, multiline, dotall, Unicode, and sticky modes. All processing runs locally in your browser using the native JavaScript regex engine — no data is sent anywhere, ever.

Create cron expressions using visual dropdowns, presets, or paste an existing expression to decode it in plain English instantly. Supports standard 5-field cron syntax (minute, hour, day of month, month, day of week) used by crontab on Linux and macOS, GitHub Actions schedule triggers, Kubernetes CronJobs, GitLab CI pipeline schedules, Jenkins, AWS EventBridge, and Google Cloud Scheduler. The built-in cheat sheet covers every special character including asterisk, slash, hyphen, and comma, plus shortcut strings like @daily, @weekly, and @monthly. Runs entirely in your browser with no server calls, no signup, and no data collection — completely private by design.

Convert text between 9 naming conventions used across programming languages, databases, APIs, URLs, and written content. Type or paste any text and instantly see it in UPPERCASE, lowercase, Title Case, Sentence case, camelCase, PascalCase, snake_case, kebab-case, and CONSTANT_CASE — all at once, updating live as you type. JavaScript uses camelCase for variables, Python and Ruby use snake_case, CSS uses kebab-case for class names, Java and React use PascalCase for classes and components, and Go uses PascalCase to export identifiers. Instead of memorizing the rules or rewriting names by hand, paste your text once and copy the convention you need. The tool splits words intelligently from any input format — including mixed-case strings, acronyms, and special characters — then reassembles them in each target case. Everything runs locally in your browser with no server, no signup, and no data collection.

Generate cryptographically secure passwords using the Web Crypto API (crypto.getRandomValues), a CSPRNG seeded by your operating system's entropy pool — the same source of randomness used by HTTPS, SSH, and TLS. NIST SP 800-63B recommends focusing on password length over complexity rules, and this tool lets you create passwords from 4 to 64 characters with full control over uppercase, lowercase, numbers, and symbols. A real-time strength indicator shows estimated entropy in bits and approximate crack time against modern GPU attacks. Keep a history of recently generated passwords and copy any of them with one click. All generation happens locally in your browser — no passwords are ever transmitted, stored, or logged on any server.

Generate realistic fake data for development, testing, and prototyping. Choose from 7 data types — person, company, address, product, user account, transaction, and blog post — and generate 1 to 100 records as JSON. Powered by @faker-js/faker, the most widely used fake data library in the JavaScript ecosystem with over 70 locale options. Faker.js provides modules for person, location, company, finance, internet, commerce, lorem, date, and many more data categories, producing output that looks convincingly real while containing zero actual personal information. All generation happens locally in your browser — no server calls, no API keys, no data leaving your device.

Type or paste text and instantly see cryptographic hashes computed using the Web Crypto API (crypto.subtle.digest), the same cryptographic engine that powers HTTPS and TLS in every modern browser. All four SHA-2 family algorithms run simultaneously as you type — SHA-256 (256-bit, the industry standard used in blockchain, digital signatures, and file integrity), SHA-384 (384-bit, common in government and financial systems), SHA-512 (512-bit, the strongest variant with superior performance on 64-bit processors), and SHA-1 (160-bit, included for legacy compatibility but deprecated by NIST since 2011 after the Google/CWI SHAttered collision attack proved it insecure). Color-coded output with bit-length display, one-click copy, and real-time avalanche-effect demonstration — change a single character and watch every hash change completely. All computation is local; your text never leaves your device.

Paste any text and get an instant probability score of whether it was written by AI or a human. Uses a RoBERTa-based classifier model trained by OpenAI, running entirely in your browser via Transformers.js. No signup, no server, no API calls. Your text stays on your device.

Paste or type any text and instantly see the exact token count for OpenAI models (GPT-4o, GPT-4.1, GPT-5, o1, o3, o4-mini) or get a close estimate for Claude, Gemini, and Llama. This free token counter uses gpt-tokenizer, a production-grade BPE tokenizer that runs entirely in your browser — your text never leaves your device. Use it to check prompt size before API calls, estimate costs, optimize prompts to fit context windows, and debug tokenization by viewing individual token IDs and decoded strings.

Compare AI model pricing across every major provider — OpenAI, Anthropic, Google, Meta, Mistral, DeepSeek, xAI, and more. Pricing is fetched live from the OpenRouter API so numbers are always current with zero maintenance. Enter your expected input and output token counts, set your number of requests, filter by provider, and instantly see what each model would cost for your workload.

Paste or type any text to instantly see all eight writing metrics at once: word count, character count with spaces, character count without spaces, sentence count, paragraph count, line count, estimated reading time (based on 200-250 WPM average), and estimated speaking time (based on 130-150 WPM presentation pace). Everything runs locally in your browser using JavaScript — no signup, no server, no data sent anywhere. Use it to hit essay word limits, stay within social media character caps, estimate how long a blog post takes to read, or time a speech before you deliver it.

Write or paste Markdown and instantly see it rendered as HTML with a live preview. Powered by marked, the most widely used open-source Markdown parser for JavaScript with millions of weekly npm downloads. Supports full GitHub Flavored Markdown (GFM) — tables, task lists, strikethrough, fenced code blocks, autolinks, and blockquotes — on top of the CommonMark specification. Switch between four layout modes: side-by-side, stacked, editor-only, and preview-only, plus fullscreen mode for distraction-free writing. All processing happens locally in your browser — no signup, no server, no data sent anywhere.

Paste raw or minified JSON and instantly format, beautify, validate, or minify it. Features line numbers, error line highlighting, JSON path display on click, configurable indentation, and one-click copy of both formatted and minified output. All processing happens locally in your browser — no signup, no server, no data sent anywhere.

Paste two versions of any text and instantly see the differences highlighted with additions, deletions, and changes. Switch between four diff modes — character, word, line, and sentence — to get exactly the level of detail you need. Powered by jsdiff, the most widely used JavaScript diff library, implementing the Myers diff algorithm (the same algorithm used by git). All processing happens locally in your browser — no signup, no server, no data sent anywhere.