Prosus has introduced ToqanClaw, a tool builder reminiscent of OpenClaw, designed for its 5 million merchants.

      The group listed in Amsterdam is showcasing conversational app-building for restaurants and retailers, claiming that AI has previously overlooked these users. Prosus has introduced ToqanClaw, a platform that enables individuals to create apps, dashboards, and automations by simply describing their needs in natural language, similar to how they would communicate a task to a coworker.

      The technology group based in Amsterdam revealed this today, positioning itself as the first company in Europe to provide OpenClaw-like tools to its business partners on a large scale. The goal is to eliminate the need for engineers in the process. A restaurant owner seeking a delivery analytics dashboard, or a retailer wanting to automate a weekly report, can articulate their requirements in a conversation and have the tool developed without having to code, open a ticket, or await assistance from an IT team.

      Prosus asserts that the platform is immediately operational and utilizes more than 20 underlying AI models to identify the most suitable one for a specific task, which it claims makes it more cost-effective than other options. ToqanClaw is based on Toqan, Prosus’s proprietary AI platform developed for internal use, which has a data approach that ensures a customer’s data remains under their control and is not utilized to train third-party models.

      This aspect is central to the entire initiative. Earlier reports indicated that Prosus was working on an OpenClaw competitor to address European privacy issues, and ToqanClaw is the result of those discussions. This stance differs from competitors like Tencent, which have developed their solutions directly on the platform with its ClawPro enterprise agents.

      The target users include those whom Prosus claims traditional AI has overlooked. The platform is being made accessible to over five million restaurants, merchants, and entrepreneurs within its ecosystem—businesses that lack the technical teams to create their own software.

      The narrative aligns with the wider democratization argument that has accompanied the rise of OpenClaw and its agent tools, now aimed more at small businesses instead of developers. Prosus backed its approach with data: Chief Executive Fabricio Bloisi mentioned that the group spent 18 months developing it internally, resulting in 60,000 agents and 10,000 applications created by individuals who had never coded before.

      The company also highlighted the training of a specialized commerce model, referred to as the Large Commerce Model, which was developed using data from over a billion customers and 500 million daily interactions. By connecting it to ToqanClaw, agents can begin to predict business needs instead of merely following instructions.

      The launch included customer examples, all provided by the company. Lebkov & Sons, a Dutch cafe chain, reportedly reduced its financial reporting time from weeks to just 30 minutes. Burger & Frites, a burger chain in Rotterdam, created a delivery analytics agent that they claim saves approximately €21,000 monthly. Poke Perfect, a poke-bowl chain, developed a WhatsApp-based operations assistant that decreased routine staff inquiries by 70%. These figures come from Prosus and have not been independently verified.

      In addition to ToqanClaw, Prosus also expanded the availability of Zapia, a consumer AI assistant it supports. Zapia manages tasks from start to finish; for instance, it can find a restaurant, shortlist options in a family group chat, wait for votes, and make reservations. Prosus reports that over six million people, mainly in Latin America, currently use it. It is available on the App Store, Google Play, and the web, with a free tier for most personal usage and a paid plan for more frequent users.

      Both launches coincided with Prosus Forward, the group’s inaugural product event, which presented an AI-first strategy emphasizing that ownership of data and customer relationships, rather than the model itself, is the real advantage. The more challenging question, as with any tool that claims to generate software from a simple sentence, is whether merchants will continue to use what they create once the demonstration concludes.