Alibaba Cloud has announced a series of initiatives designed to strengthen the capabilities of developers in creating artificial intelligence applications on a global scale.
Dongliang Guo, Vice President of International Business and Head of International Products and Solutions at Alibaba Cloud Intelligence, stated, "Alibaba Cloud is committed to delivering real value to global developers through cutting-edge AI models, enhanced cloud infrastructure, and accessible support programs. Together, we aim to spark more AI-driven innovations, benefiting startups, enterprises, and industries altogether across the globe."
The company has introduced its latest proprietary large language models under the Qwen family, including the Qwen2.5 series, offering models with parameter sizes from 7 billion to 72 billion. These models are accessible to developers globally via APIs on Alibaba Cloud's Model Studio, a generative AI development platform. The platform also includes multimodal AI models, such as the vision understanding models Qwen-VL series, visual generation model Wanx2.1, and the audio language model Qwen-Audio.
Alibaba Cloud has developed a proprietary AI coding assistant, Tongyi Lingma, which utilises the Qwen 2.5-coder model. This assistant supports developers in tasks such as code completion, optimisation, debugging assistance, code snippet search, and batch unit test generation.
Several new AI development tools are now available on Model Studio. These include Workflow for enhancing task breakdown and workflow control, and Agent for multi-agent collaboration. Other features include Retrieval-Augmented Generation (RAG) for enhancing generative AI accuracy, Batch Reasoning for simultaneous response generation, Automated Model Evaluation (AutoEval), and services for model deployment and application observability. These tools aim to provide developers with a comprehensive suite for improving productivity and creativity.
To support critical workloads, Alibaba Cloud is set to launch its 9th Generation Enterprise Elastic Compute Service (ECS) instances in April 2025. These instances offer a 20% improvement in computing efficiency compared to previous models. Furthermore, network performance enhancements through eRDMA (elastic Remote Direct Memory Access) can potentially increase support for high-performance computing by up to 50%.
Alibaba Cloud's new Container Compute Service (ACS) is now available to international customers. This service aims to streamline workload deployment using container technology, integrating container services with Alibaba Cloud's underlying resources to reduce costs and manage infrastructure more effectively.
The company has also launched the Alibaba Cloud GenAI Empowerment Program, which aims to foster development and innovation among global developers and startups using the Qwen models. Participants can expect free cloud credits, access to training workshops, invitations to tech shows, and opportunities for co-marketing efforts.
A Japanese company, Axcxept, has utilised the Qwen 2.5 large language model to develop an open-source AI model called EZO. This model excels in tasks such as coding, information extraction, math reasoning, role-playing, and writing in Japanese. EZO outperforms existing models and is intended for applications in fields like healthcare and public institutions.
Kazuya Hodatsu, CEO of Axcxept Inc, commented, "Qwen 2.5 has significantly enhanced its ability to process Japanese, giving it a competitive edge over other models. With Axcxept's proprietary training process, we have developed a Japanese LLM that delivers unmatched accuracy."
Another user of the Qwen-based models is OxValue.AI, a venture from the University of Oxford, which integrates these models for AI-driven company valuation services. This involves processing text and audio data to deliver precise and economical valuation assessments to corporate clients.
Professor Xiaolan Fu, Founder of OxValue.AI, said, "Processing diverse data sources is essential for our valuation services. With the support of Alibaba Cloud's AI models, we've significantly improved the quality and efficiency of this process. By collaborating with Alibaba Cloud, we're able to deliver greater value to our corporate clients."
