Skip to content
RentThe6ix
Shortlist

About the data

Every number on RentThe6ix comes from public data and our own code. Gemini reads your search and writes plain-language explanations, but it never produces a rent estimate, a travel time, a score or a ranking. AI interprets; code calculates.

Datasets

Each dataset below was loaded by a script that recorded its source, licence and the date we downloaded it.

Map: base map tiles from OpenFreeMap, built from OpenStreetMap data (ODbL 1.0), credited in the corner of every map.

Where the listings come from

We collect rental listings from the public pages of these sites. Each listing links back to its original post, where you contact the landlord.

  • Kijiji5,655 listings shown · last seen Oct 17, 2026
  • Realtor.ca4,306 listings shown · last seen Oct 18, 2026
  • TorontoRentals4,075 listings shown · last seen Oct 18, 2026
  • Rentals.ca2,522 listings shown · last seen Oct 18, 2026
  • Zillow1,096 listings shown · last seen Oct 18, 2026
  • We store a short excerpt of each description (280 characters at most), with phone numbers, emails and names removed. Contact details are never stored.
  • Photos are never downloaded or stored: the cover photo is shown straight from the source site.
  • The same unit listed on several sites is shown once, as its cheapest copy, unless that copy is more than 15% below the others (often a copied ad). The other copies and their prices appear under “Also on”.
  • A listing not seen for 8 days is marked stale, and after 21 days it is removed from search.
  • If a site blocks or rate-limits our collector, it stops for that run. We never solve captchas or use proxies.

How the rent check works

We estimate what a listing should rent for from similar listings nearby that are still on the market. These are asking rents, not signed leases.

  • Comparables have the same number of bedrooms (rooms are compared only with rooms), at most 1 bathroom more or less, and were seen in the last 60 days.
  • We search within 1 km, 2 km, 3 km, widening until there are at least 5. With fewer, the listing shows “Not enough data”.
  • Closer, more recent and more similar listings count more. One building counts as at most 2 comparables, so a single tower can't set the price.
  • Comparables are adjusted to the listing's terms: utilities included $150, parking $100, furnished $150, in-suite laundry $50 a month.
  • The estimate is the weighted median. A listing more than 10% below it is “Underpriced”, more than 10% above is “Overpriced” (15% when confidence is low), and in between “Reasonably priced”.
  • More than 30% below the estimate is flagged “Unusually low for this area”: check that the listing is real before sending money.
  • The Statistics Canada Toronto-area average asking rent is shown for context. It never changes the estimate.

How the match score works

Listings that fail one of your must-haves are left out. The rest get a match score from 0 to 100, made of five parts:

  • Value (30%, or 45% if you ask for a good deal): the rent against our estimate.
  • Price (15%): how the rent sits against your budget, when the budget is a preference.
  • Travel (30%): time to your places against your target.
  • Home (25%): how the unit matches your wishes: bedrooms, size, AC, laundry, pets and so on.
  • Reported crime levels (off, or 15% if you turn them on): the neighbourhood's crime rate against the city's.
  • A detail the listing doesn't state scores 50, halfway between a match and a miss. Unknown is never treated as “no”.
  • A part with no data is left out and the others are scaled up, so the parts always add up to the total shown.
  • Listings that confirm every must-have rank above ones that leave some unconfirmed. Ties go to higher price-check confidence, then more confirmed details, then the most recently seen.

Travel times

Travel times are estimates from the distance between the listing and the place, not routed trips:

  • Transit: 18 km/h along streets about 1.4× the straight-line distance, plus 8 min for waiting, parking or getting to the stop
  • Walk: 4.8 km/h along streets about 1.25× the straight-line distance
  • Bike: 15 km/h along streets about 1.25× the straight-line distance, plus 1 min for waiting, parking or getting to the stop
  • Car: 25 km/h along streets about 1.35× the straight-line distance, plus 3 min for waiting, parking or getting to the stop

Real trips vary with routes, transfers, schedules and traffic. Driving times assume no traffic.

Reported Crime Levels

Reported crime levels are off unless you turn them on. They use the Toronto Police Service's police-reported rates for five major crime types in the listing's neighbourhood, compared with the city average. Reported crime isn't all crime, rates move from year to year, and they say nothing about any street or building. We never call an area safe or unsafe.

How we use AI

  • Gemini turns your search into filters you can see and edit, and each one shows the words it came from.
  • Gemini writes the short explanation of why a listing ranks where it does, from numbers our code computed. Explanations that mention a number are checked against those numbers.
  • In our collector, Gemini reads listing descriptions to fill details the site doesn't list (utilities, laundry, parking and so on), and each value must quote the phrase it came from.
  • What we send: your search text, and listing descriptions with contact details already removed. No account or personal data is sent. If Gemini is unavailable, a simpler non-AI version is used.

Limitations

  • Rents are asking rents from listings, not what tenants actually pay.
  • Locations come from the listing; an address we couldn't match to the City's records is shown as approximate.
  • Listing sites differ in what they state, so many details are unknown. We mark them as unknown rather than guessing.
  • Neighbourhood facts come from the 2021 Census and other dated sources; each shows its period.
  • Building scores cover RentSafeTO's buildings only: apartment buildings with 3 or more storeys and 10 or more units.