Price tracker features for every store you shop at

A product is one thing you want to buy; links are the stores that sell it.

• Add a product with links from up to 3 online stores (25 products per account on the hosted service).
• The first price captured is the base price; the cheapest in-stock link is the best price; the all-time low is kept separately.
• Check interval per product, as often as every 6 hours, with random jitter, exponential back-off on failures and a polite delay per store.
• Pause, archive, tag and add notes to products, and set a target price.
• Each store’s price is converted into the product’s currency with daily European Central Bank exchange rates.

Prices are read in a fixed order and the first reliable result wins.

• Built-in adapters for Amazon, eMAG, eBay, Altex/Media Galaxy, cel.ro, PC Garage, evoMAG and a few more.
• schema.org JSON-LD (Product, Offer, AggregateOffer) and microdata, then Open Graph and product meta tags.
• Price heuristics that score price-like elements and ignore old, crossed-out and related-product prices.
• An optional AI fallback that reads the page text with your own API key when everything else fails.
• Understands formats such as 1.299,99 lei, 1,299.99, 1 299 zł and £51.77, plus stock phrases in English, Romanian, German, French, Spanish, Italian, Polish, Czech and Hungarian.

The main alert fires when a link drops and is now cheaper than every other store. You can add more rules per product.

• Cheapest store: a store became the cheapest for this product.
• Below target, percentage below base price and back in stock.
• Any price change (opt-in), link keeps failing and research finished.
• Cool-down and de-duplication, quiet hours, an alert inbox in the app and a delivery log for every alert.

Connect up to 3 channels on the hosted service and limit each one to the alert types you care about.

• Discord: a webhook URL; alerts arrive as embeds with the product image.
• Slack: an incoming webhook.
• Telegram: a bot token and chat ID.
• ntfy: a topic on ntfy.sh or your own ntfy server; important alerts get a higher priority.
• Webhooks: any URL, POST or PUT, custom headers and an optional HMAC-SHA256 signature.
• Use Test to send a sample. Email alerts are not available on the hosted service yet.

See whether today’s price is actually a good one.

• Price history chart for every store link, with 7-day, 30-day, 90-day, one-year and all-time ranges.
• Base price, best price now, all-time low, highest price, 30-day average and low.
• You save: the difference between the best price now and the base price.
• A dashboard with savings across products, recent checks and failing links.

AI is optional and runs on your own API key. Everything else works without it.

• Research with AI (from the Add button): the model plans search queries, PriceSentinel opens each result, reads the real price and asks the model whether it is the exact same product. You tick the ones to track.
• AI assistant: chat with your price data, for example what dropped this week?, and let it add links or change check intervals.
• Providers: Anthropic, OpenAI, OpenRouter, Groq, Google Gemini (OpenAI-compatible), Mistral, DeepSeek or any OpenAI-compatible endpoint.
• Keys are encrypted at rest. Scheduled AI extraction uses your own tokens.
• Hosted limits: 10 research jobs and 50 assistant messages per day.

The app works in any modern browser, on desktop and on your phone.

• Dashboard, product pages, alert inbox and activity log with live updates.
• Light and dark theme and a mobile layout.
• Import and export (JSON backup or CSV), and account deletion from the settings.
• No ads, no analytics, no trackers. Only a session cookie.

The open-source version runs the same tracker on your own hardware, without limits.

• Email alerts through your own SMTP server, and Gotify.
• Built-in MCP server so AI assistants can read prices and alerts.
• Local AI models through Ollama or LM Studio, and private web search with SearXNG.
• Optional headless Chromium for shops that render prices with JavaScript, domain rules and an HTTP proxy.
• REST API for everything the app does (API reference).