Elastic announced on its official website the acquisition of Jina AI. This acquisition marks Elastic's further expansion in the open-source retrieval and multimodal AI fields. Xiao Han, the former CEO of Jina AI, will serve as Elastic's VP of AI, responsible for AI strategy and R&D. The core team led by Xiao Han will continue to focus on research in vector models, re-rankers, Readers, and small models.
Image source note: The image was generated by AI, and the image licensing service provider is Midjourney
Jina AI was founded in 2020 and has raised approximately $37.2 million in funding, with investors including Canaan Partners, Mango Capital, GGV Capital, SAP.iO, and Yunque Capital. As an active open-source community, Jina AI focuses on multimodal embeddings, small models, and retrieval re-ranking algorithms, becoming an important innovator in vector search and semantic retrieval. This acquisition means that Jina AI's technology and capabilities will enter a broader enterprise application scenario, helping Elastic enhance its product competitiveness.
After the acquisition announcement, Ken Exner, CPO of Elastic, stated that retrieval technology is the core of the Elastic Search AI platform. To enhance this advantage, Elastic's collaboration with Jina AI will focus on providing strong search relevance, helping users find and analyze the required information in real-time and accurately. Jina AI has always been a pioneer in the open-source multimodal, multilingual vectors, re-rankers, and small language models, especially in combining large language models (LLMs) with precise context and proprietary data. Jina AI's experience will be beneficial for enterprises in building top-tier retrieval solutions.
Jina AI's capabilities include general vector models that can deeply understand text and images, supporting single-vector and multi-vector outputs; advanced re-rankers that perform excellently in multilingual long documents, code search, and visual document retrieval tasks; and small language models that can complete high-quality tasks such as converting HTML to Markdown and extracting data from HTML to JSON.
In the future, Elastic will continue to provide Jina AI's models for free on Hugging Face, and through Elastic Cloud's Elastic Inference service, achieve native integration, allowing customers to directly call relevant features within the Elastic vector search environment. Xiao Han expressed his anticipation, emphasizing that combining Jina AI's models with Elastic's ecosystem will help developers create faster and more accurate AI experiences.
Jina AI will continue to uphold the spirit of open source, focusing on high-quality search models, accelerating the realization of its mission through Elastic's larger platform. This acquisition is not the end of Jina AI, but the beginning of something new. In the future, the Jina AI team looks forward to working with more developers to build the next generation of open-source AI search.
Key points:
🌟 Elastic announced the acquisition of Jina AI, promoting the integration of open-source retrieval and multimodal AI technologies.
🤝 Former CEO of Jina AI, Xiao Han, will serve as VP of AI at Elastic, continuing to drive related technology development.
🚀 Jina AI's technology will be widely applied on the Elastic platform, improving the quality of enterprise search solutions.