REKA

Reka AI is a technology company based in Sunnyvale, California, specialising in AI research and product development. Founded by Cyprien de Masson d’Autume, Dani Yogatama, Mikel Artetxe, Qi Liu, and Yi Tay, Reka AI focuses on creating multimodal AI models that can interpret and generate content across various data types, including text, images, audio, and video. The company offers solutions for industries like media, entertainment, security, defence, and content creation.
 

Reka’s platform allows users to perform a wide range of tasks, such as searching through images and videos, summarising long-form content, and detecting complex events for real-time alerts. The company’s flagship product, Reka Vision, is designed to work across video and image content, enabling advanced searching, reasoning, and summarisation at scale. Reka’s models, such as Spark, Flash, and Core, offer different levels of performance tailored to specific use cases, from embedded devices to enterprise-level applications.
 

One of Reka’s key offerings is its AI-driven tools for creators and enterprises. For creators, Reka provides tools to generate video reels, highlight sports content, and repurpose old footage. Enterprises can use Reka’s models for security applications, such as video analysis for incident detection, or for enhancing business intelligence through advanced research tools. The company’s products are powered by novel multimodal transformers, which are built from scratch to solve complex problems and provide reliable, efficient solutions.
 

Reka AI is also committed to open-sourcing its technology and providing enterprise-level deployments. The company continues to innovate in the AI space, offering tailored solutions that integrate seamlessly with a variety of business needs. It is known for its flexibility, collaborative work environment, and global reach, with a team spread across multiple time zones.

Reka AI was founded in 2022 by Dani Yogatama, Cyprien de Masson d’Autume, Qi Liu, Mikel Artetxe, and Yi Tay, all of whom previously worked at leading AI research institutions such as DeepMind, Meta FAIR, and Google. 

The company was established with the aim of developing multimodal AI models that could efficiently process and understand text, images, audio, and video. Recognising the limitations of large language models (LLMs) in handling diverse tasks, the founders sought to create specialised models that could perform specific functions more effectively.
 

In June 2023, Reka AI emerged from stealth mode and raised $58 million in a funding round led by DST Global Partners and Radical Ventures, with participation from Snowflake Ventures. This funding enabled the company to accelerate its research and development efforts. 

Later that year, in July, Reka secured an additional $50 million in funding, bringing its total to $108 million. This round was led by DST Global Partners and Snowflake, with the valuation estimated at approximately $300 million. The company also announced a partnership with Snowflake to integrate its models within Snowflake's platform, allowing users to run third-party models like Reka's while ensuring data privacy.
 

In July 2025, Reka AI raised $110 million in a funding round backed by Nvidia and Snowflake, bringing its valuation to $1 billion. This funding supported the development of Reka's multimodal models, including Reka Flash, a 21-billion-parameter model capable of understanding video, images, text, and audio. The company also introduced Reka Vision and Reka Research, platforms designed for visual search and complex question answering, respectively.
 

Today, Reka AI continues to advance in the field of multimodal AI, offering solutions that cater to various industries, including media, entertainment, security, and enterprise applications. The company's products are utilised by organisations such as Shutterstock and Turing Video, and it has established partnerships with government agencies to provide AI solutions for public safety and efficiency. Reka AI remains committed to developing efficient, high-performing AI models that integrate seamlessly into diverse workflows, aiming to empower enterprises and creators worldwide.

Reka AI's mission is to build advanced AI models that improve human lives and drive innovation across various sectors. The company aims to develop multimodal AI systems capable of understanding and generating text, images, audio, and video. By focusing on efficiency and customisation, Reka strives to provide businesses and creators with powerful, easy-to-use tools. Their work helps industries like media, security, and content creation to automate tasks, enhance productivity, and unlock new possibilities in AI technology. Reka is committed to making cutting-edge AI accessible, driving scientific progress and enabling companies to solve complex problems more efficiently.

Reka AI envisions a future where multimodal AI becomes integral to every business, creator, and researcher. The company aims to bridge the gap between advanced AI research and real-world applications. Reka’s vision is to build AI models that work seamlessly across text, video, images, and audio, providing intelligent solutions to a wide range of industries. The company strives to empower its users by making powerful AI tools accessible, customisable, and efficient. Reka believes that its innovations can help businesses operate smarter, enhance creativity, and solve complex problems, driving technological progress worldwide.

Reka AI has garnered recognition for its innovative approach in developing multimodal AI models. In 2023, the company raised significant funding, achieving a $300 million valuation. Their platforms, such as Reka Vision and Reka Research, have received praise for revolutionising industries like media, security, and content creation. With their advanced models integrated into platforms like Snowflake, Reka continues to build partnerships with leading organisations. As a company focused on excellence and research-driven solutions, Reka AI has been recognised as a promising leader in AI, gaining credibility among investors, industry leaders, and global enterprises.

Reka AI offers a range of innovative products and services designed to advance the capabilities of artificial intelligence across multiple industries. The company focuses on developing multimodal AI models that can understand and generate content from text, images, audio, and video. These models are tailored to meet the needs of businesses, creators, and developers, enabling them to streamline processes, enhance productivity, and unlock new possibilities in AI.
 

One of Reka AI's flagship products is Reka Vision. This tool is built to process and analyse video and image content at scale. It allows users to search through millions of images or video files using natural language, summarise long videos, and detect complex events in footage to trigger real-time alerts. It is particularly useful for industries like media, entertainment, and security, as it can quickly analyse vast amounts of visual data, saving time and increasing efficiency. Reka Vision is ideal for tasks such as content research, archive repurposing, and even performance analysis in sports and media production. It can automatically create highlight reels, summarise key events, and track player performance over time.
 

Another key product is Reka Research, which is an advanced agentic AI model designed for complex question answering and research tasks. Reka Research can browse the web and private documents, synthesising information from multiple sources to provide in-depth answers to research questions. It eliminates the need for manual data gathering and processing, completing tasks that would typically take hours in just a few minutes. The product is integrated with Google Drive, allowing businesses to access internal knowledge and combine it with external sources. It also provides transparent agentic reasoning, meaning every action taken by the system is traceable, with citations for the sources used. This makes Reka Research a valuable tool for market intelligence, academic research, and internal operations.

For content creators, Reka AI offers a powerful tool to generate social media reels and highlights. Reka for Creators helps users turn long-form content, such as YouTube videos or podcasts, into short, engaging clips ready for sharing on platforms like Instagram and TikTok. The tool uses Reka’s multimodal AI to automatically identify interesting moments in videos, add captions, and even generate relevant hashtags. It supports customisation, allowing creators to adjust the length, aspect ratio, and metadata to suit their needs. This product helps creators save time while maximising the reach and engagement of their content.
 

In addition to these core products, Reka AI provides enterprise-ready solutions tailored to specific industries. These include services for media and entertainment, where Reka AI's models can be used for content research, creation, and editing. For example, the platform can automatically generate product highlights, create social media ads, and analyse large video archives. Reka's Archive Repurposing feature allows companies to search decades of video content, tagging and categorising old assets for licensing or remixing.
 

For security and defence sectors, Reka offers advanced video analysis and surveillance tools. These can detect behaviours like fighting, loitering, or theft in real-time, helping security teams respond faster and more accurately. The system can also search through traffic incidents and generate incident reports based on natural language inputs. This functionality is highly beneficial for smart city projects and law enforcement agencies looking to enhance their surveillance capabilities.
 

Lastly, Reka AI provides solutions for research engagement. This tool helps businesses and organisations understand which topics resonate with their target audiences across social media and other platforms. It can also analyse the performance of specific campaigns, helping users optimise their content strategies.