Crawl and extract clean data for LLM training, RAG applications, and AI model development.
Sorted by popularity. Each actor is tested weekly. Click any tool to see detailed features, pricing, and user reviews.
Web search and fetch tool for AI agents and RAG pipelines. It queries Google Search, scrapes the top N pages using a full web browser, and returns their content as clean Markdown for further processing by an LLM. Can also fetch individual URLs.
Crawl websites and extract text content to feed AI models, LLM applications, vector databases, or RAG pipelines. The Actor supports rich formatting using Markdown, cleans the HTML, downloads files, and integrates well with 🦜🔗 LangChain, LlamaIndex, and the wider LLM ecosystem.
AI-powered manga page colorizer. Takes black-and-white manga image URLs and returns fully colored pages with the original line art preserved. Supports batch processing and multiple original languages for accurate text region handling.
Transcribes audio files from URL or upload into text with timestamps, language, and confidence. Supports WAV, MP3, AIFF, AAC, OGG, FLAC, and more.
Extracts text from manga page images using OCR and translates it into one or more target languages. Returns original text and translations in a flat structured dataset.
Crawls websites from starting URLs and returns each page's full text content as Markdown and cleaned HTML, ready for AI and LLM pipelines.
Upload a resume and a job description to get back an optimized CV tailored to the target role. Controls for tone, format, and creativity level.
Generate pixel art images from text descriptions. Choose style, size, and color palette. Get base64-encoded PNGs in a structured dataset.
Generates thumbnail images from a video file or a text topic using Gemini AI. Returns each thumbnail as an image file in the chosen aspect ratio.
Extract structured JSON data from PDF files automatically. Upload one or more PDFs and get clean, organized data with AI-powered field extraction.
Converts HTML from URLs, pasted content, or uploaded files into structured JSON using OpenAI. Specify fields or let AI auto-detect important data. Returns one flat JSON object per input.
Converts uploaded images or image URLs into black and white coloring pages with bold outlines and simplified backgrounds. Adjust AI temperature for precise or artistic results.
Extract public Sherdog data: UFC and MMA fighter records with method breakdowns, full bout-by-bout fight history, and entire event cards for finished and upcoming shows.
Scrapes Hugging Face model cards by task, library, license, language, or keyword search. Returns each model as a flat row with downloads, likes, author, tags, and last-modified date.
Scrapes historical OHLCV price data from Stooq for a single ticker, index, or forex pair. Returns each bar as a flat row with date, open, high, low, close, and volume.
Restores vintage and damaged photos from image URLs using AI. Fixes tears, scratches, fading, and stains, and optionally colorizes black and white images.
Enhance restaurant photos with AI: improve lighting, color, and detail. Upload URLs, choose enhancement level, and export professional images.
Normalizes addresses from a list or CSV file into standardized components like street, city, state, ZIP, and country. Supports custom normalization rules.
Upload product photos or portraits and turn them into professional studio-style renders. Pick from preset styles, backgrounds, and lighting setups, or write your own custom descriptions for full creative control.
Fetches web pages from search queries or direct URLs and returns clean content as markdown, text, and optional HTML. Designed to feed RAG pipelines and LLMs with structured data.
Analyzes candidate applications for fraud signals and returns a 1-10 authenticity score with plain-language explanations. Feed it resumes, cover letters, work histories, or a URL to a candidate feed.
Scrapes the OpenRouter model catalog and returns each model as a flat row with prompt price, completion price, context length, modality, and provider. Filter by provider, modality, price ceiling, and context window.
Extract US and Canadian Thoroughbred racing data from Equibase: race entries with jockey, trainer, owner, sire and dam, finished results with win-place-show and exotic payoffs, and horse profiles with career records.
Scrapes Understat player season stats for Premier League, La Liga, Bundesliga, Serie A, Ligue 1, and RFPL. Returns each player as a flat row with xG, xA, shots, key passes, and per-90 rates.
Scrape Civitai AI model cards by search query, type, and base model. Returns creator info, stats, version details, trigger words, and download URLs as a flat dataset.
Scrapes Kaggle dataset listings by search term, tag, file type, license, or size. Returns one flat row per dataset with title, URL, size, votes, and optional full description and version history.
Upload ID document images or PDFs and get structured JSON back. Extracts fields like firstName, dateOfBirth, and documentNumber from passports, driver's licenses, and identity cards using AI. No OCR setup required.
Scrapes the Australian TGA ARTG register by keyword and product type. Returns each therapeutic good entry with its ARTG ID, product name, active ingredients, and sponsor as a flat row.
Scrapes Start.gg tournament listings filtered by game, country, and upcoming status. Returns each tournament as a flat row with name, date, location, and game details.
Power ratings, 1,300+ team and player stat leaderboards, betting odds, projections and situational trends for NFL, NBA, MLB, WNBA and college sports.
Scrape VegasInsider odds for NFL, NBA, MLB, NHL, WNBA and college. Spreads, totals and moneylines from 9 sportsbooks, with opening lines, line movement, ticket and handle splits, futures, player props and scores.
Scrapes documentation sets from DevDocs by name or keyword. Returns each docset as a flat row with name, slug, and version. Export to CSV, JSON, Excel, or XML.
Queries the public Nationalize.io API for a list of first names and returns each country probability prediction as a flat row with the name, country code, probability, and sample count.
Scrapes AI model metadata from Replicate's Explore page or any collection. Returns each model as a flat row with name, description, run count, owner, and hardware details.
Scrapes the Shanghai ARWU university ranking for a given year and returns each university as a flat row with world rank, national rank, total score, and indicator values.
Scrapes superinvestor portfolio holdings from Dataroma, including stock positions, allocation percentages, and recent buy/sell activity. Returns one flat row per holding.
Scrapes live cryptocurrency market data from Nomics. Returns price, market cap, volume, and supply metrics for each coin in a flat row. Supports USD, EUR, BTC, and ETH quote currencies.
Scrapes FutureLearn course listings by search term or from the full catalog. Returns each course as a flat row with title, provider, URL, and category. Supports up to one million courses per run.
Scrapes Brawl Stars brawlers, maps, and game modes from Brawlify. Returns each item as a flat row with attributes like rarity, class, star powers, gadgets, map environment, and mode rules.
Scrapes NASA JPL Horizons ephemeris data for a chosen body, observer center, and time span. Returns each sampled epoch as a flat row with position, velocity, and physical data.
Scrapes NCI Thesaurus cancer ontology concepts by search term or NCIt code, returning synonyms, definitions, semantic types, and parent-child hierarchies as flat rows.
Scrapes curated drug annotations from MyChem.info for a search query or a list of drug names or InChIKeys. Returns each drug as one flat row with indications, mechanisms, targets, and identifiers.
Scrapes ModelScope public model listings from a search query. Returns each model as a flat row with downloads, stars, task, and organization metadata.
Scrapes confirmed exoplanet data from the NASA Exoplanet Archive, filtering by discovery method, host star, and discovery year, and returns each planet as one row.
Scrapes Reactome pathway records by search term, stable ID, or species, and returns each record as a flat row with stable ID, name, type, and species.
Scrapes asteroid and comet records from the NASA JPL Small-Body Database by designation or catalogue browse. Returns each object as a flat row with orbital and physical fields.
Scrapes geologic unit records from Macrostrat by latitude/longitude or stratigraphic name. Returns each unit as a flat row with attributes like name, age, and lithology.
Scrapes Orphanet rare-disease records by ORPHAcode or from the latest Orphadata list. Returns each disease as a flat row with term, synonyms, definition and epidemiological data.
Scrapes Cleveland Museum of Art open access collection. Returns artwork title, artist, date, medium, and image URL. Filter by keyword, department, or object type.
Extract embedded JSON-LD structured data from Wikipedia pages. Returns schema.org fields including @type, name, description, author, publisher, date, license, and interaction statistics. No API key required.
Scrapes live ECB and quarterly US Treasury exchange rates. Returns each currency pair as a flat row with the date and source, ready for financial models and reports.
Scrapes clinical trial records from ClinicalTrials.gov, WHO ICTRP, OpenFDA, and ClinVar. Returns each trial as a flat row with registry ID, phase, status, enrollment, and eligibility criteria.
Scrapes Hugging Face Collections by search term, owner, or sort order. Returns each collection as a flat row with title, description, upvotes, and model references.
Scrapes U.S. government publication metadata from GovInfo by collection, date range, and optional Congress or document class filters. Returns each publication as a flat row with its title, URL, and modification date.
Scrapes the Rijksmuseum's online collection by search term, object type, or maker name. Returns each artwork as a flat row with title, maker, date, medium, and link.
Scrapes the Hugging Face Open LLM Leaderboard for model names, benchmark scores (ARC, HellaSwag, MMLU, TruthfulQA, Winogrande, GSM8K), and parameter counts. Returns one flat row per model.
Scrape SportyTrader odds comparison, betting tips, live scores, lineups, match stats and bookmaker reviews. 13 bookmakers, 15 football markets, 18 sports. No login, no API key, no proxy.
Scrape valued images from any Wikimedia Commons category via the MediaWiki API. Returns page ID, namespace, and title for each image. No login required.
Scrapes license template metadata from Wikimedia Commons categories using the MediaWiki API. Returns each template as a flat row with pageid, namespace, and title. Supports any category and up to one million items per run.
Scrape Covers.com odds for 12 leagues from up to nine sportsbooks. Public betting percentages, line movement history, player prop projections with EV, injury reports, futures, and computed hold, fair value and arbitrage. No proxy.
The full list, sorted by user count. The most popular ai & llm actors are shown as cards above.
Pick a tool from our ai & llm collection. Most have free tiers. Get your first data in under 5 minutes.