Milestone achieved just seven months after official launch, averaging 4.8 new services per day
95% of partners continue operations, emerging as a “gateway to entrepreneurship” for developers

Toss announced that the number of partner mini apps on its mini app platform “Apps in Toss” has surpassed 1,000.
The milestone comes just seven months after the platform’s official launch in July, and about ten months after the first mini debuted within the Toss app. Since the official launch, an average of 4.8 new mini apps have been added to the Toss ecosystem each day.
The early growth of Apps in Toss was led by the gaming sector. The convenience of instant execution, without requiring a separate app download, has resonated strongly with game developers, fueling a string of success stories. One notable example is Manababa, the developer of the game Doldoldi. Facing financial difficulties and even considering closing its office at the time, the company rebounded after partnering with Apps in Toss, surpassing KRW 210 million in monthly revenue and gaining new momentum for growth.
The non-gaming sector is also expanding rapidly. Non-game mini apps now account for more than 60% of the total, spanning a wide range of categories including lifestyle, health, and AI. Notable examples include an indie developer who launched 20 services using “vibe coding” technology, which enables code implementation through natural language descriptions, as well as a solo developer who released five mini apps within a single month and generated KRW 25 million in monthly revenue. These cases highlight Apps in Toss as a viable gateway to entrepreneurship.
Beyond simply expanding its partner base, the platform has also demonstrated meaningful results in terms of service sustainability. Over the past ten months, 95% of partners that joined Apps in Toss are still actively operating their services. This strong retention rate underscores the platform’s ability to address one of the biggest hurdles in app development and entrepreneurship - acquiring an initial user base - while providing a stable foundation for growth.
“In the case of the recently viral ‘Dubai Chewy Cookie Map,’ everything from partnership discussions to mini app development and launch was completed in just one day,” said a Toss representative. “We will continue building an ecosystem where anyone with an idea can quickly and easily launch an app, helping foster an environment where entrepreneurship is more accessible than ever.”


Milestone achieved just seven months after official launch, averaging 4.8 new services per day
95% of partners continue operations, emerging as a “gateway to entrepreneurship” for 

Toss Surpasses 1,000 Partner Mini Apps on Its “Apps in Toss” Platform

Presents large-scale federated learning technology that optimizes AI model training in environments where personal information protection regulations prevent data from being centralized on a server
The research shows privacy and model performance can be achieved simultaneously

Toss announced today that a research paper authored by one of its employees has been accepted to the Neural Information Processing Systems (NeurIPS), one of the world’s most prestigious academic conferences in artificial intelligence. The study was conducted by Jinu Lee, Machine Learning Engineer of the Toss Face Modeling Team, in collaboration with Seoul National University’s Computer Vision Lab.
NeurIPS is widely regarded as the most influential conference in machine learning and neural information processing, with a paper acceptance rate of only around 20 percent. This year’s conference is taking place from December 2 to 7 at the San Diego Convention Center in the U.S., bringing together researchers from around the world to share the latest AI advancements.
The accepted paper titled “FedLPA: Local Prior Alignment for Heterogeneous Federated Generalized Category Discovery” introduces Federated Local Prior Alignment (FedLPA), a Federated Learning technology that optimizes AI model training in environments where personal information protection regulations prevent data from being centralized on a server. It addresses a key limitation of conventional federated methods, whose performance degrades significantly when data characteristics vary widely across clients* or user groups, or when encountering entirely new data types.
The research team combined Infomap-based local clustering—which automatically groups data with similar characteristics across clients* or user segments—with a Local Prior Alignment technique that stabilizes training through prediction alignment. This approach allows each device to autonomously understand and leverage its own data distributions, demonstrating strong performance in Generalized Category Discovery (GCD) by accurately identifying novel categories even when category types or data distributions are unknown in advance.
The study further shows that privacy and model performance can be achieved simultaneously: FedLPA enables the development of global AI models that comply with the legal requirements of countries with strict data-protection regulations.
“This research focuses on optimizing federated learning to remain effective even in highly constrained environments where data cannot be moved to a server due to regulatory constraints, each client* has a different data distribution, and the number of new categories is completely unknown,” said Jinu Lee, Machine Learning Engineer at Toss.
A Toss spokesperson added, “It is a significant milestone that Toss’s AI capabilities have been formally recognized for the first time by a world-class academic conference. We will continue advancing practical, service-ready technologies to deliver more sophisticated AI-powered services while upholding strong privacy protection.”

*A separate data source or device that has its own data and its own AI model.



Presents large-scale federated learning technology that optimizes AI model training in environments where personal information protection regulations prevent data from being centra

The research shows privacy and model performance can be achieved simultaneously

Toss Paper on AI Model Training Optimization Accepted at Top Global Conference

New personalized services for small business owners and individuals unveiled
Hands-on demo zones set up for FacePay and the soon-to-launch Healthy Living Score


At Korea Fintech Week 2025, Toss CEO SG Lee (left) introduces Toss payment terminals and the FacePay service to FSC Chairman Eog-weon Lee

Toss is participating in Korea Fintech Week 2025, where it is presenting its facial-recognition payment service FacePay, alongside a suite of new AI- and data-driven services.
Hosted by the Financial Services Commission (FSC), the three-day event runs from November 26 to 28 at the aT Center in Seoul under the theme “Fintech × AI, The Personalization of Finance.” Leveraging its AI-based data analytics capabilities, Toss is highlighting inclusive financial services designed for both small business owners and individual users.
At its café-themed booth, Toss is showcasing offline payment technologies - including Toss FacePay, Toss Place payment terminals, and POS system - alongside six upcoming AI- and data-powered services developed to meet the needs of small merchants and consumers. They are:
Healthy Living Score, which analyzes financial and lifestyle data to predict how daily habits may influence insurance benefits
Toss Score, an alternative credit scoring model designed for individuals with limited credit history
Mobility Living Score, which uses in-app activity and other data to predict the likelihood of car accidents
Toss Analytics, which visualizes commercial district trends nearby by analyzing sales flows and customer foot traffic
Toss SOHO Score, which evaluates creditworthiness based on real cash flow and business performance
Business Loan Refinance Brokerage Service, which compares loan terms across multiple financial institutions and helps users switch to more favorable options
“Our booth is designed to show how technology-driven fintech can be applied in everyday life,” said a Toss representative. “We will continue advancing AI- and data-driven innovations to deliver a more convenient financial experience for both small business owners and consumers.”


New personalized services for small business owners and individuals unveiled
Hands-on demo zones set up for FacePay and the soon-to-launch Healthy Living Score


At Korea Fintech W

Hands-on demo zones set up for FacePay and the soon-to-launch Healthy Living Score

Toss Showcases AI-Powered Inclusive Finance Services at Korea Fintech Week 2025

1,127 people registered for the event
22 sessions delivered across three tracks

Toss announced today that it has successfully concluded its annual security conference, GUARDIANS 2025.
Now in its third year, GUARDIANS brings together security experts from the Toss community to share hands-on experience and best practices, and to discuss the latest issues shaping today’s security landscape. This year’s event more than doubled in size compared to last year and was open to anyone interested in security. A total of 1,127 participants registered in advance.
The conference opened with a keynote by Toss CISO Jungho Chi, who emphasized that “keeping pace with evolving security threats requires new approaches in addition to traditional methods.,” adding, “I hope the real-world security cases shared today offer practical insights for those working in the field.”
A total of 23 security specialists from the Toss community took the stage as speakers, delivering 22 sessions organized across three parallel tracks. Participants were free to choose from a broad range of topics, including personal data protection, threat response, asset protection, and applications of AI in information security.
“This conference provided an opportunity for those working in corporate security to share their challenges and explore solutions together,” said a Toss spokesperson. “We will continue to develop GUARDIANS as a platform that contributes to the advancement of Korea’s information security industry.”


1,127 people registered for the event
22 sessions delivered across three tracks

Toss announced today that it has successfully concluded its annual security conference, GUARDIANS 2

1,127 people registered for the event, 22 sessions delivered across three tracks

Toss Successfully Concludes Security Conference ‘GUARDIANS 2025’

Adds an average of 8,200 new users per day
Only facial-recognition payment service to receive preliminary adequacy review approval from Korea Personal Information Protection Commission.

Toss announced that FacePay, its new offline payment service powered by facial-recognition technology, has surpassed one million users.
FacePay allows users to complete payments simply by looking at the terminal - no credit card or phone required. With sub-one-second transaction speed, high convenience and advanced security, the service is rapidly gaining traction. After users register their face and  payment method in the Toss app, they can use FacePay at any businesses equipped with Toss terminals.
The service officially launched in September, about six months after undergoing beta testing that applied facial recognition technology to existing Toss POS terminals installed at select locations including convenience stores.  Since launch, the user base has grown from 400,000 to 1 million - an average of 8,200 new users per day, or roughly one new user every 10.5 seconds.
The service is also seeing strong on-site usage. As of November, nearly half of all FacePay-enabled merchants nationwide recorded at least one transaction. In Seoul, that figure reached about 79%, and usage was particularly high among cafés, with around 95% of participating locations recording actual payments.
Toss plans to accelerate the expansion of FacePay, supported by its differentiated security framework. FacePay is the only facial-recognition payment service in Korea to receive preliminary adequacy review approval from Korea Personal Information Protection Commission. The service incorporates multiple layers of AI-based protection to detect and mitigate security risks, including liveness anti-spoofing, a Fraud Detection System (FDS), and an advanced facial recognition model capable of precisely distinguishing facial variations and similar faces.
“Beyond convenience and speed, the trust we have built through our long-standing focus on security has been the foundation for FacePay’s rapid adoption,” said a Toss representative. “We will continue expanding supported locations and features so that more users can experience the ease and delight of seamless payments in everyday life.”


Adds an average of 8,200 new users per day
Only facial-recognition payment service to receive preliminary adequacy review approval from Korea Personal Information Protection Commis

# press release

Toss Surpasses 1,000 Partner Mini Apps on Its “Apps in Toss” Platform

Toss Paper on AI Model Training Optimization Accepted at Top Global Conference

Toss Showcases AI-Powered Inclusive Finance Services at Korea Fintech Week 2025

Toss Successfully Concludes Security Conference ‘GUARDIANS 2025’

Toss’ FacePay Tops 1 Million Users Two Months after Official Launch