New wave of AI tools from China

Source: DeepSeek. Benchmarks for DeepSeek-R1 compared to OpenAI's o1.

DeepSeek has released an open-source AI reasoning model that it said is on par with the OpenAI's most advanced o1, which is proprietary. Prior to the DeepSeek launch many thought that it would take more time to match OpenAI's lead in AI. Another disruptive difference vs market leader OpenAI is that the DeepSeek-R1 release can be used for free as DeepThink on its website and can be commercialised freely, said the company.  

DeepSeek also added six small models based on Qwen and Llama into the open source pool, and said that its 32B and 70B models are on par with OpenAI's o1-mini. Alibaba Cloud also has a number of open source AI models in Github under its Qwen label.

Alibaba Cloud, the digital technology and intelligence backbone of Alibaba Group, separately unveiled an expanded suite of large language models and AI development tools, upgraded infrastructure offerings, and new support programmes for global developers at its recent annual developer summit. The advancements aim to empower developers worldwide to build innovative AI applications more cost-effectively and drive a thriving global generative AI community.

“Alibaba Cloud is committed to delivering real value to global developers through cutting-edge AI models, enhanced cloud infrastructure, and accessible support programmes,” said Dongliang Guo, VP of International Business, Head of International Products and Solutions, Alibaba Cloud Intelligence. 

“Together, we aim to spark more AI-driven innovations, benefiting startups, enterprises, and industries altogether across the globe.” 

The latest Qwen models, Alibaba Cloud’s proprietary large language model family, including the Qwen2.5 series with sizes ranging from 7 billion to 72 billion parameters, are now globally accessible via APIs on its generative AI development platform, Model Studio. Additionally, multimodal AI models, including vision understanding models such as the Qwen-VL series, visual generation model Wanx2.1 (also known as Tongyi Wanxiang), and audio language model Qwen-Audio, are also available to developers. 

Developers can further leverage Tongyi Lingma, Alibaba Cloud’s proprietary AI coding assistant powered by the Qwen 2.5-coder model. The AI programmer offers features such as code completion and optimisation, debugging assistance, code snippet search and batch unit test generation. 

In addition to a broader range of models, new AI development tools are also accessible by global developers on Model Studio. These tools include Workflow, which breaks down complex tasks into subtasks to enhance workflow control, and Agent, which supports multi-agent collaboration for planning and execution tasks. 

Other tools such as RAG (retrieval-augmented generation), which helps enhance the accuracy and reliability of generative AI models with external sources; Batch Reasoning, which generates responses simultaneously with multiple prompt inputs; AutoEval (automated model evaluation), as well as model deployment and application observability services will be available by end-January. 

To facilitate AI and other critical workloads, Alibaba Cloud revealed that its 9th Generation Enterprise Elastic Compute Service (ECS) instance will be available in global markets starting April 2025. The latest generation of ECS instances has better performance enhancements compared to its previous iteration, including a 20% increase in computing efficiency. Additionally, by accelerating networks through eRDMA (elastic remote direct memory access), its performance in supporting high-performance computing, search recommendations, and Redis databases can be further improved by up to 50%. Redis offers in-memory data storage.

The Alibaba Cloud Container Compute Service (ACS) is now available for international customers. Designed for simplified and optimised workload deployment using container technology, the ACS integrates container services with underlying cloud computing resources, significantly reducing costs and technical complexity. 

To foster innovation, Alibaba Cloud introduced the Alibaba Cloud GenAI Empowerment Program, an accelerator programme for global developers and startups leveraging its Qwen models to build generative AI applications. Participants can gain support, including free cloud credits, training workshops, invitations to tech shows and demo days, as well as product co-marketing opportunities. 

The new Alibaba Cloud Generative AI Solution Whitepaper was also released. Developers can find useful resources such as the latest trends and use cases for generative AI, as well as more about Alibaba Cloud’s generative AI products. 

Axcxept, a Japanese company specialising in AI products such as voice assistants, has developed an open-source, lightweight AI model called EZO based on the Qwen 2.5 large language model (LLM). EZO outperforms state-of-the-art models in areas such as coding, information extraction, math, reasoning, roleplay, and writing in Japanese. With low latency and robust performance, EZO is tailored to serve industries such as healthcare and public institutions in Japan, ensuring secure and efficient AI applications, Alibaba Cloud said. 

“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,” said Kazuya Hodatsu, CEO, Axcxept. 

OxValue.AI, a deep-tech venture from the University of Oxford, uses Alibaba Cloud’s Qwen-based multimodal AI models for AI-driven company valuation services. By processing and analysing text and audio data related to financing, R&D, and operations, OxValue achieves precise and cost-efficient valuation assessments tailored to corporate clients.

“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 can deliver greater value to our corporate clients,” said Professor Xiaolan Fu, Founder of OxValue.AI.