On Tuesday, tech giant Microsoft unveiled its latest innovation: a lightweight artificial intelligence model dubbed Phi-3-mini. This move signals the company’s strategic efforts to cater to a broader clientele by offering more affordable options in the rapidly evolving landscape of AI technology.
Phi-3-mini marks the initial release among a trio of small language models (SLMs) by Microsoft. The company is betting big on these models, recognising their potential to revolutionise various industries and redefine how people engage with technology in their professional lives.
Speaking on the cost-effectiveness of Phi-3-mini, Sébastien Bubeck, Microsoft’s vice president of GenAI research, highlighted its significant price advantage over comparable models in the market. “Phi-3 is not slightly cheaper, it’s dramatically cheaper,” Bubeck stated, noting a cost reduction of up to tenfold compared to its competitors with similar capabilities.
Designed to handle simpler tasks, SLMs like Phi-3-mini offer practical solutions tailored for companies operating with limited resources. This strategic focus aligns with Microsoft’s commitment to democratise AI and make it more accessible to businesses of all sizes.
Phi-3-mini is now readily available on Microsoft’s cloud service platform Azure’s AI model catalog, as well as on machine learning model platform Hugging Face and Ollama, a framework for local machine model deployment. Moreover, the SLM is optimised for Nvidia’s graphics processing units (GPUs) and integrated with Nvidia’s software tool Nvidia Inference Microservices (NIM), further enhancing its accessibility and performance.
Just last week, Microsoft injected $1.5 billion into G42, an AI firm based in the UAE. Additionally, Microsoft has forged strategic partnerships with innovative startups like Mistral AI, facilitating the integration of cutting-edge AI models into its Azure cloud computing platform.