What you get
Four ways into the same corpus. Three of them need no key and no code.
See plans and pricing API reference
Explore
Pick a slice of the corpus and see what is in it. Choose a technology, a product type, a country, a launch type and a date range. It tells you what those sites are built with, which domain endings they use, what kind of products they are, and whether the slice is growing week on week. A short example list sits underneath, and the same filters carry across to a CSV when you want the rows themselves.
It aggregates over exactly the population the export would return, so the shape you see on screen and the rows you download cannot disagree.
Looking is free, and re-opening a slice you have already run is free. What is metered is the number of new slices we calculate for you in each billing period.
Stackdar alerts
Standing watches over newly detected launches. "Tell me when a new launch ships with Supabase", or "alert me on every discovered launch in Germany". Delivery is by email digest, Slack webhook, or HMAC-signed webhook.
Alerts fire only on real launch-time detections, never on backfills, so an alert always means a new launch shipped with the technology, not that we newly learned to detect it.
You can narrow a watch by technology, category, country and launch type, by several technologies at once, and by technologies that must be absent.
A watch can cover startup launches only, or every new site we find. Every new site is several times the volume of startup launches alone. The form tells you what the watch would have sent over the last thirty days before you save it.
Email digests go out at an hour you choose, in your own timezone, and keep the same local time through daylight saving. You can take the list as a CSV attachment rather than a long email. Webhooks are near-instant and are the right channel for anything high volume.
The every-launch stream is a watch with no technology filter. You name nothing to follow and it delivers every new launch we detect. It is only available on Firehose, and goes by webhook or Slack, never email. Tens of thousands of sites a day is not something to put in an inbox.
Downloads and extra site data
Downloads from your account. Pick a technology, category, country, launch type and date range on the download page, and get a spreadsheet back. Each download costs one credit, and every launch in the file counts once against your distinct-launch allowance. A search that matches nothing is not charged for.
More data on the site. While you are signed in, the public technology and category pages show more: more launches on each page, every tool in a category instead of the first few, and the full list of technologies that appear alongside the one you are looking at. Bulk downloads stay on the download page or in the API, where they are metered.
The API
Give it a domain and get the launch and its stack back: the launch record, the technologies detected on its most recent crawl with per-technology confidence and first and last seen dates, and parsed email posture. If we do not know the domain, the API returns not found.
You can also query the corpus as lists: a newest-first feed filterable by technology, category, country, launch type, indie flag and launch-date window, plus per-technology launch lists, weekly adoption trends across the full history, and CSV export of any list query.
Some rows carry more than the stack: hosting and infrastructure, a quality score, and how ready the site is for AI agents. Those come from the deeper crawl, which not every site gets. The count is in the numbers below, so you can see how much of the corpus that covers before you buy.
The feed, per-technology lists and trends start at the Pro plan. Domain lookup, CSV export and Stackdar watches are on every plan, so an Indie plan covers watching your own corner of the market, while Pro lets you query the wider corpus.
Every endpoint, every parameter, error behaviour and quota headers are in the API reference.
You can also connect an AI assistant instead of writing code. Paste https://mcp.stackscope.dev into Claude, ChatGPT, Cursor or any MCP client and ask in plain language. It is on every plan and uses the same permissions and allowance as the API. See Connect your assistant.