Jina AI

Jina AI is a company developing an open-source platform for building multimodal artificial intelligence (AI) services and pipelines. Jina offers products for three main markets:

• Power users—multimodal tools for enhanced productivity
• Developers—an open-source multimodal AI stack
• Enterprises—multimodal AI strategies for different business needs

Jina is a Python platform for serving machine learning models from local deployment to advanced orchestration frameworks, such as Docker-Compose, Kubernetes, and Jina AI Cloud. Jina aims to allow users to focus on logic and algorithms rather than infrastructure complexity. With Jina users can perform the following:

• Build and serve models for any data type and any mainstream deep learning framework
• Design high-performance services, with scaling, duplex client-server streaming, batching, dynamic batching, async/non-blocking data processing, and any protocol
• Serve LLM models while streaming their output
• Utilize docker container integration via Executor Hub, OpenTelemetry/Prometheus observability
• Host streamlined CPU/GPU services via Jina AI Cloud
• Deploy to their own cloud or system with Kubernetes and Docker Compose integration

Headquartered in Berlin, Jina AI was founded in February 2020 by Han Xiao (CEO), Nan Wang, and Bing He. The company raised $37.5 million in its first twenty months, including a $30 million Series A round in November 2021 led by Canaan Partners. Other participants included Mango Capital and previous investors GGV Capital, SAP.iO, and Yunqi Partners. As of 2023, the company has grown to over 250,000 users and 40,000 developers worldwide with locations now in San Jose, Beijing, and Shenzhen, as well as Berlin.

In October 2023, Jina launched its second-generation text embedding model jina-embeddings-v2, an open-source model that supports 8K (8192 tokens) context length.