Helsingborg’s conversation AI startup Klang raises €1.32 million and releases open speech-to-text model for Swedish

Klang, a Helsingborg-based AI company that transcribes, summarises and analyses interviews, meetings and dialogues, has raised €1.32 million (SEK 15 million) in fresh capital at a valuation of €13.2 million (SEK 150 million).

The round was backed by existing owners and led by Johan Lenander, entrepreneur and tech investor, together with Emil Sjödin, founder of Refined, and Daniel Gadd, co-founder of Position Green, among others.

“In a world where AI performs more and more tasks, human conversation becomes an increasingly important source of information. At the same time, today’s technology for understanding conversations is still limited. Our vision is that what is said should never be lost. We want to create a perfect memory of every conversation – a memory that not only remembers what was said, but also understands the conversation as well as, or even better than, the people in the room,” said Niklas Silfverström, co-founder and CEO, Klang.

Founded in 2023 by Silfverström, Viktor Silfverström, Andreas Argelius, Mattias Fält, and Jacob Andersson, Klang develops AI solutions for transcribing, summarising, and analysing conversations. 

The company states that this year it has focused on a large-scale investment in its own speech and language models, built to handle real-time conversations with speed and accuracy. It is built on scalable technology that can be launched quickly in more markets.

It recently released its first open speech-to-text model for Swedish. The model, called Pianissimo, is trained for Swedish and Swedish dialects and can run locally on a standard laptop. It is released as open source, freely available for others to use and build on.

Klang notes that Pianissimo builds on Nvidia Parakeet, an open speech-to-text model developed for fast, efficient transcription, which it has optimised for Swedish and Swedish dialects. It reported that, in its tests, the model achieved accuracy comparable to KB Whisper Large, the best open Swedish model from the National Library of Sweden, but was roughly 60 times faster.

“We’re building more and more technology that needs real-time transcription ourselves, but we’ve lacked an open model adapted for Swedish that combines high quality with the speed we need. The fact that Pianissimo runs on ordinary consumer hardware also opens up entirely new kinds of services. Now we’re looking forward to seeing what others build on the model,” said Mattias Fält, Head of AI, co-founder and lead for Pianissimo at Klang AI.

Klang’s long-term vision is to create a platform that serves as a complete, searchable memory of the conversations of companies and organisations, where AI understands not only what was said but also the context, nuance, dialects, and how different conversations connect.

Klang highlighted that it is developed and operated in Sweden, with its own models adapted to European data protection. 

“AI is changing not only the product we build but also how we build the company. A small, technically strong team can today achieve what used to require a much larger organisation. Now we above all want to pick up the pace in research, product development and distribution,” Silfverström added. 

The company plans to use this capital to fund continued AI development and international expansion. It plans to recruit around 10 people in areas such as research, development, and marketing. In the near term, Klang is focusing on expansion in the Nordics, Benelux, the DACH region and Japan.

On 12 August 2026, Klang launched in Japan. Users can record, transcribe and summarise conversations through a Japanese interface. Both the web platform and the iOS and Android apps are localised for the market.

The post Helsingborg’s conversation AI startup Klang raises €1.32 million and releases open speech-to-text model for Swedish appeared first on EU-Startups.