International apps in the ChatGPT store
A French classifieds marketplace has the most parameters of any app in the entire ChatGPT App Store. leboncoin ships 184 parameters across its tools, more than Monday.com (159), Canva (122), or Stripe (57). When we analyzed all 147 third-party apps in the store, we found that the international apps aren't just present; they're building some of the most technically ambitious integrations in the ecosystem.
We identified 16 apps that are clearly non-US or non-English-first brands, spanning eight distinct markets. Here's what we found.
The International Roster
| App | Country/Region | Category | Params | % Required | Enum Params |
|---|---|---|---|---|---|
| leboncoin | France | SHOPPING | 184 | 3.26% | 21 |
| 요기요 | South Korea | FOOD | 33 | 39.39% | 12 |
| DoneDeal | Ireland | SHOPPING | 29 | 3.45% | 3 |
| realestate.com.au | Australia | LIFESTYLE | 26 | 7.69% | 3 |
| komoot | Germany | LIFESTYLE | 24 | 8.33% | 0 |
| Trade Me Property | New Zealand | LIFESTYLE | 19 | 0% | 0 |
| Cafe24 | South Korea | SHOPPING | 16 | 25% | 6 |
| Kleinanzeigen | Germany | SHOPPING | 13 | 7.69% | 0 |
| CHECK24 | Germany | TRAVEL | 8 | 100% | 0 |
| Jobkorea | South Korea | PRODUCTIVITY | 8 | 100% | 0 |
| DEWA | UAE (Dubai) | BUSINESS | 6 | 33.33% | 0 |
| LOTTE CHEMICAL | South Korea | BUSINESS | 3 | 100% | 0 |
| ImmoScout24 | Germany | LIFESTYLE | 2 | 100% | 0 |
| hipages | Australia | LIFESTYLE | 2 | 100% | 0 |
| freee確定申告 | Japan | FINANCE | 1 | 100% | 0 |
| AutoScout24 | Germany | LIFESTYLE | 1 | 100% | 0 |
That's 16 apps across eight countries and territories. Not a massive share of the 147-app store, but enough to reveal some clear patterns.
Germany Leads the International Pack
The German-speaking market has the strongest international presence with five apps: ImmoScout24, AutoScout24, Kleinanzeigen, CHECK24, and komoot. Add leboncoin (French, but operating in a similar European marketplace tradition) and you have a cluster of six European marketplace and comparison-shopping apps.
What's striking is the type of company showing up. They're established, dominant marketplace brands:
- ImmoScout24 is Germany's largest real estate platform
- AutoScout24 is one of Europe's biggest auto marketplaces
- Kleinanzeigen (formerly eBay Kleinanzeigen) is Germany's leading classifieds site
- CHECK24 is a major comparison shopping platform for insurance, travel, and utilities
- komoot is a popular outdoor navigation app with a strong European user base
Germany's over-representation probably reflects a few things: the country's high digital literacy, a marketplace-heavy consumer economy, and a tech ecosystem that's large enough to have engineering teams capable of building ChatGPT app integrations. Germany is Europe's largest economy and has long been one of the strongest adopters of e-commerce platforms, so the fact that German marketplace brands are showing up early in the ChatGPT App Store is less surprising than it might initially seem.
Korea's Focused Approach
South Korea contributes four apps: 요기요 (food delivery), Jobkorea (job search), Cafe24 (e-commerce), and LOTTE CHEMICAL (industrial/B2B). This is a diverse spread across consumer and enterprise categories, which suggests that ChatGPT platform interest in Korea isn't limited to one sector.
요기요 is particularly interesting from a technical standpoint. It has 33 parameters and 12 enum parameters, putting it in the top tier for enum usage alongside apps like Autotrader (13 enums) and Cloudinary (11 enums). Those enum params include things like sort, servingType, and category, which are the kind of structured, filterable search parameters you'd expect from a food delivery app that needs to help users narrow down restaurants by cuisine type, delivery method, and sorting preference.
Cafe24, a Korean e-commerce platform, takes a similar approach with 16 parameters and 6 enum params covering things like shop_currency_code, sort, order, and color. It's the kind of integration designed for merchants who want to query their store data through ChatGPT.
The Parameter Design Split
One of the more interesting patterns we noticed is that international apps seem to split into two distinct design philosophies.
The "flexible search" model: leboncoin (184 params, 3.26% required), DoneDeal (29 params, 3.45% required), Kleinanzeigen (13 params, 7.69% required), realestate.com.au (26 params, 7.69% required), and komoot (24 params, 8.33% required) all ship with very low required-parameter rates. They're exposing a wide surface area of optional filters, letting ChatGPT (and by extension the user) choose how narrowly to scope a query. This is a classifieds-marketplace pattern: you want to let users search by price, location, condition, category, or any combination, without requiring all of them.
The "structured query" model: CHECK24 (8 params, 100% required), Jobkorea (8 params, 100% required), ImmoScout24 (2 params, 100% required), hipages (2 params, 100% required), and LOTTE CHEMICAL (3 params, 100% required) go the other direction, making every parameter required. These apps want tightly scoped, specific queries rather than broad browsing.
For context, the store-wide average is 34.58% required parameters. The international marketplace apps pulling that number down are making a deliberate architectural choice: they're building for broad, exploratory search rather than narrow, transactional queries.
leboncoin: The Parameter Outlier
leboncoin deserves its own section because it's genuinely unusual. At 184 total parameters, it's the most parameter-heavy app in the entire ChatGPT App Store, ahead of Atlassian Rovo (171), Monday.com (159), and Canva (122). Only 6 of those 184 parameters are required (3.26%), which means the app exposes 178 optional filters.
It also has 21 enum parameters, tying it with Atlassian Rovo for the third-highest enum count in the store (behind Klaviyo at 32 and Monday.com at 30). Those enums cover vehicle-specific attributes like fuel, gearbox, doors, seats, and vehicle_type, plus general marketplace filters like condition, seller_type, and sort_by.
This makes sense when you consider what leboncoin actually is: France's largest classifieds platform, covering everything from used cars to apartments to electronics. To build a useful ChatGPT integration, you need to support searching across multiple product verticals, each with its own set of relevant attributes. A car search needs fuel type and gearbox filters. A phone search needs brand and memory capacity. A general search just needs a keyword. leboncoin's approach is to expose all of those as optional parameters and let the model figure out which ones are relevant for a given query.
Whether this approach works well in practice (does ChatGPT actually handle 184 optional parameters gracefully?) is an open question we'd love to test.
The Pacific Rim: Australia and New Zealand
Three apps come from Australia and New Zealand: realestate.com.au, hipages, and Trade Me Property. All three are in the LIFESTYLE category, and all three are property or home services platforms. This is a narrow but logical cluster: real estate is inherently local, and these are brands that dominate their domestic markets but have no reason to expand to the US.
realestate.com.au is notable for shipping just 1 tool (property.search_for_sale) but 26 parameters, including 3 enum params (constructionStatus, userQueryCategory, sortBy). It's a single-tool app that packs a lot of search flexibility into that one tool.
Trade Me Property, from New Zealand, takes the most relaxed approach of any international app: 19 parameters with 0% required. Every single parameter is optional, which means you could technically call its tool with no arguments at all.
What's Missing
The absence of certain markets is as telling as the presence of others. There are no apps from:
- China (where the AI ecosystem is largely separate)
- India (despite its massive developer population and English-language tech sector)
- Latin America (no Brazilian, Mexican, or Argentine apps)
- Southeast Asia (no apps from Indonesia, Vietnam, Thailand, or the Philippines)
- Africa (no representation from any African market)
Some of these gaps are explainable. China has its own AI platforms and app ecosystems. But the absence of Indian apps is more surprising, given that India has a large, English-speaking tech workforce and many consumer platforms (like Swiggy, Zomato, or Flipkart) that could plausibly build ChatGPT integrations.
The current international mix skews heavily toward wealthy, digitally mature markets in Europe, East Asia, and the Pacific. Whether this reflects OpenAI's go-to-market priorities, local developer interest, or just the early stage of the platform is hard to say.
What This Tells Us About Global Adoption
A few takeaways from the international app data:
Marketplace and classifieds platforms are the early movers. The most common pattern among international apps is "search our listings through ChatGPT." This makes sense: marketplace search is a natural fit for conversational AI, and these companies already have structured product data ready to be queried.
International apps aren't building differently. Despite coming from different markets and engineering cultures, the international apps follow the same design patterns as US apps. They use the same parameter types, the same required/optional splits, and the same tool structures. The ChatGPT app platform is creating a common technical language across markets.
Domestic market leaders are the ones showing up. Every international app we identified is a market leader or major player in its home country. These aren't startups experimenting with ChatGPT; they're established companies making a strategic bet on the platform. That's a signal worth paying attention to.
The German cluster is the one to watch. Five apps from a single market (six if you include France) suggests something more systematic than individual company decisions. If German marketplaces are collectively deciding that ChatGPT integration is worth the investment, it may indicate that the platform is gaining traction with European consumers in ways that aren't yet visible in US-centric coverage.
Looking ahead, we're curious whether the next wave of international apps will come from the same markets (deepening the German and Korean presence) or from new regions entirely. We're also interested in whether these international apps see meaningful usage, or whether the ChatGPT user base is still too US-centric for a French classifieds platform or a Korean food delivery app to get much traction. As the store grows beyond 147 apps, the international mix will be one of the more revealing indicators of where ChatGPT is becoming a genuine consumer platform rather than a Silicon Valley curiosity.
Methodology
This analysis covers 147 third-party apps in the ChatGPT App Store as of February 2025. We excluded integrations built and maintained by OpenAI (like GitHub, Linear, Slack, and Google Workspace) to focus on apps that companies built and shipped independently.
Want access to the full dataset? Contact us to learn more.