I built a page that shows the inflation rate for every country, and the first thing I did after publishing it was compare Venezuela against Trading Economics. My page said 252%. Theirs said 534.2%. Both numbers were real. That gap ended up shaping most of the work, so this post is mostly about it.
The page lives at khaledalam.net/inflation-rate. It covers about 200 countries, goes back to 1960, and it's free, with no sign-up and no ads.
Why build another inflation tracker?#
Most of the good ones sit on proprietary data. You can look, but you can't download anything or check where a number came from, and some of them hide the history behind a paywall.
Meanwhile the IMF and the World Bank publish all of this for free. Nobody had put it together in a way I actually wanted to use. So I did it myself, using only open, official sources, with every number traceable back to where it came from.
What can you do with it?#
The main view is a table you can sort, filter by continent, or search, plus a world map coloured by the latest figure. Click a country and you get its own page:
- the latest monthly figure and the previous one
- a chart of the last 36 months
- the full history since 1960, with the IMF's projections drawn dashed
- how it compares with its region and the world
- average inflation for each decade
- inflation next to GDP growth and unemployment
- a purchasing-power calculator
Every country has its own link, like /inflation-rate/#EGY for Egypt, so you can send someone straight to it.
So why was Venezuela wrong?#
It wasn't wrong exactly. It was answering a different question.
My first version used annual averages from the IMF World Economic Outlook. For Venezuela in 2025 that's 252%: prices across the whole of 2025 compared with the whole of 2024. Trading Economics shows the latest month against the same month a year earlier. In August 2026 that was 534.2%.
When people say "inflation is X%", they almost always mean the second one. An annual average is the right number for history and long-run comparisons, but it's a year old by the time it settles. So the page now shows both: the latest monthly figure in the table, and the annual series for the history charts.
Where do the numbers come from?#
No single free source has a recent monthly figure for every country, so the page pulls from several and keeps whichever has the newest one:
- the IMF's CPI dataset, which covers most countries
- the OECD
- the BIS (Bank for International Settlements)
- the World Bank's Global Economic Monitor
- Banco Central de Venezuela, for Venezuela
Venezuela is the reason that last one is there. It isn't in any of the international monthly datasets. The central bank does publish its consumer price index as a spreadsheet, though, and computing the yearly change from it gives exactly 534.2% for August 2026. That's the same figure Trading Economics shows, which was a nice moment.
The history comes from the IMF World Economic Outlook from 1980 onward, and from the World Bank before that. Forecasts are the IMF's own.
Right now 177 of the 201 countries have a figure from the last year. The rest are mostly countries at war, like Sudan, Yemen and Syria, plus a few tiny territories. Nobody publishes recent monthly data for them, so the page shows the latest annual figure and labels it as an average.
What went wrong along the way?#
A few things I didn't expect:
- APIs are picky about who's asking. One API rejected my requests until I changed the User-Agent header. Another rejected them until I changed it back. They want opposite things.
- Some "monthly" data is quarterly. New Zealand only publishes CPI every quarter, but one dataset repeats each quarter's number across three months. The page showed "Previous (May): 4.1%, Latest (June): 4.1%", which looks fine and means nothing. It now detects that pattern and shows Q2 2026 instead.
- Some years just don't exist. The IMF has no Argentina figures for 2014 to 2016, years when its official statistics weren't trusted. I left the gap empty. The chart has a break there, and the calculator won't compute across it. That felt better than inventing a number.
- Hyperinflation breaks charts. Venezuela hit 65,374% in 2018. On a normal axis, forty years of history becomes a flat line with one spike. For countries like that, the history chart switches to a log scale and says so in the caption.
What does 100 pounds from 2005 buy today?#
This is the calculator, and it's the part I keep playing with. Take Egypt. Add up each year's inflation from 2005 to 2025 and prices rose about 11.9 times. So 100 pounds in 2005 bought the same as 1,186 pounds in 2025.
Turn it around and 100 pounds kept in a drawer since 2005 buys what 8.4 pounds did back then. That's the poster at the top of this post. The notes shrink as 100, 59.6, 37.8, 19.1, 8.4, at five-year steps, and it's real data, not an illustration.
It works for any country and any pair of years with data. If you pick future years it uses the IMF projections and tells you so.
Frequently asked questions#
Why do your numbers differ from other sites?#
Usually because of what's being measured. The "Last" column is the year-on-year change for the latest month, like most news reports use. The history and averages use annual averages. Different sites also update on different days, so for a week or two after a new release one site can be ahead of another.
How often is it updated?#
It checks for new figures every day by itself. The "last updated" date only changes when a number actually changes.
Can I use the data?#
Yes. There's a JSON download link at the bottom of the page. The underlying data belongs to the IMF, OECD, BIS, World Bank and Banco Central de Venezuela, so credit them if you publish anything built on it.
Is it accurate?#
The numbers are exactly what those institutions publish. I didn't adjust, smooth or fill anything in. Where data is missing, the page shows it as missing.
If you find a country that looks off, tell me. That's how the Venezuela fix happened. I'm @khaledalam on GitHub.
Comments
No comments yet — be the first.