API reference

Three endpoints, no key, one attribution rule

The Tomorrow catalog is published as open data. Everything on screen is readable by machine over plain HTTP — no account, no token, no rate-limit header to negotiate. In exchange we ask for two things: cache what you pull, and credit the source when you publish it.

Endpoints

GET/api/public/catalog.jsonapplication/json · cache 30 min

Every tracked product and segment with version, entry price, model stack and change counts.

GET/api/public/catalog.csvtext/csv · cache 30 min

One row per product, same fields as the JSON. Opens directly in a spreadsheet.

GET/feed.xmlapplication/rss+xml · cache live

Every recorded version, price and model change as it lands. Per-product feeds at /tools/{slug}/feed.xml.

GET/api/public/tokenomics.jsonapplication/json · cache 30 min

Today's blended rate per model tier, the frontier price floor since 2020, and the everyday-task cost basket. CORS-open, built for live cost meters in other apps.

All responses send Access-Control-Allow-Origin: *, so they can be fetched straight from a browser app. Requests are logged in aggregate (path, size, user agent) to measure which agents read us. No personal data is stored.

Tell us who you are (optional)

The data stays free and keyless. But if you're a university, a lab, a newsroom or a company building on it, register a key and send it as X-Api-Key (or ?key=). It costs you nothing and it means we can show that real institutions rely on this — and warn you before anything breaking ships. We use your work email only to confirm the organisation; the address is never published or sold.

Registered callers are tracked by organisation name in our internal usage reporting. Unregistered traffic is still counted, just anonymously.

Fair use

The catalog is regenerated once a day. Polling more than a few times an hour returns the same bytes — respect the Cache-Control header instead.
Pull the whole snapshot once rather than fetching per product in a loop. There is no per-record endpoint by design.
Send a user agent that identifies you. Anonymous high-volume clients are the first thing we throttle if we ever have to.
No authentication and no service guarantee: endpoints, field names and shapes may change as the catalog grows. Breaking changes get called out in the changes feed.

Licence and terms

Free academic tier. Universities, research labs, libraries and non-profits get unlimited access, free, permanently — no quota, no approval process, no invoice later. Register with your institutional address (.edu, .ac.uk, .edu.au, .org and equivalents are detected automatically) and tell us in one line what you're building, so we know what to keep stable for you.

The dataset is licensed Creative Commons Attribution 4.0. Commercial use, redistribution and model training are all permitted.
Attribution: credit Tomorrow and link to https://tomorrow.aliensquad.ai wherever the data is displayed or quoted.
Figures are compiled from public vendor pages and archives. Usage and scale numbers are reported claims with an as-of date, not measurements — verify against the vendor before relying on them commercially.
Provided as-is, without warranty. We are not liable for decisions made on the data, and nothing here implies endorsement by the vendors tracked.
Don't present the data as your own primary research, and don't imply Tomorrow verified a claim it labels inferred or unknown.

How to cite this page

Free to cite and reuse under CC BY 4.0. Permalink: https://tomorrow.aliensquad.ai/api-docs

APA

Tomorrow. (2026). Tomorrow AI & Emerging Tech Tracker API [Data set]. AlienSquad. Retrieved 2026-09-14, from https://tomorrow.aliensquad.ai/api-docs

BibTeX

@misc{tomorrow-api,
  author       = {{Tomorrow}},
  title        = {Tomorrow AI & Emerging Tech Tracker API},
  year         = {2026},
  publisher    = {AlienSquad},
  howpublished = {\url{https://tomorrow.aliensquad.ai/api-docs}},
  note         = {Accessed: 2026-09-14}
}

Field definitions

slug: Stable identifier for the product. Safe to use as a join key; it does not change.
name / vendor: Product name and the company that ships it.
segment: Segment slug the product sits in (see the segments array).
version: The version the vendor currently presents as shipped — not the newest announced.
price_floor_usd: Cheapest paid monthly tier in USD. null means free-only or quote-only, never 'unknown'.
price_note: Verbatim pricing wording from the vendor, including the billing unit.
primary_model: Foundation model the product runs on by default, as stated by the vendor.
last_changed_at: When we recorded the most recent change (ISO date).
change_count_30d: Number of recorded changes in the trailing 30 days.
origin_region: Where the vendor is headquartered.

Downloads & citation

Grab the raw files, see the corpus size and copy a formal citation line.

How numbers are verified

The daily crawl, the archive reconciliation and the multi-model audit panel.