{"id":33425,"date":"2024-04-23T12:30:00","date_gmt":"2024-04-23T19:30:00","guid":{"rendered":"https:\/\/azure.microsoft.com\/en-us\/blog\/?p=33425"},"modified":"2025-01-31T12:07:38","modified_gmt":"2025-01-31T20:07:38","slug":"introducing-phi-3-redefining-whats-possible-with-slms","status":"publish","type":"post","link":"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/","title":{"rendered":"Introducing Phi-3: Redefining what&#8217;s possible with SLMs"},"content":{"rendered":"\n<p class=\"wp-block-paragraph\"><em>Read more announcements about Phi-3 at Microsoft Build 2024:&nbsp;<a href=\"https:\/\/azure.microsoft.com\/en-us\/blog\/new-models-added-to-the-phi-3-family-available-on-microsoft-azure\/\" target=\"_blank\" rel=\"noreferrer noopener\">New models added to the Phi-3 family, available on Microsoft Azure<\/a><\/em>.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">We are excited to introduce Phi-3, a family of open AI models developed by Microsoft. Phi-3 models are the <strong>most capable and cost-effective small language models (SLMs)<\/strong> available, outperforming models of the same size and next size up across a variety of language, reasoning, coding, and math benchmarks. This release expands the selection of high-quality models for customers, offering more practical choices as they compose and build generative AI applications.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Starting today, <strong>Phi-3-mini<\/strong>, a 3.8B language model is available on <a href=\"https:\/\/aka.ms\/phi3-azure-ai\" target=\"_blank\" rel=\"noreferrer noopener\">Microsoft Azure AI Studio<\/a>, <a href=\"https:\/\/huggingface.co\/collections\/microsoft\/phi-3-6626e15e9585a200d2d761e3\" target=\"_blank\" rel=\"noreferrer noopener\">Hugging Face<\/a>, and <a href=\"https:\/\/ollama.com\/library\/phi3\" target=\"_blank\" rel=\"noreferrer noopener\">Ollama<\/a>.&nbsp;<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li class=\"wp-block-list-item\">Phi-3-mini is available in two context-length variants\u20144K and 128K tokens. It is the first model in its class to support a context window of up to 128K tokens, with little impact on quality.<\/li>\n\n\n\n<li class=\"wp-block-list-item\">It is instruction-tuned, meaning that it\u2019s trained to follow different types of instructions reflecting how people normally communicate. This ensures the model is ready to use out-of-the-box.<\/li>\n\n\n\n<li class=\"wp-block-list-item\">It is available on Azure AI to take advantage of the deploy-eval-finetune toolchain, and is available on Ollama for developers to run locally on their laptops.<\/li>\n\n\n\n<li class=\"wp-block-list-item\">It has been optimized for <a href=\"http:\/\/aka.ms\/phi3-optimizations\" target=\"_blank\" rel=\"noreferrer noopener\">ONNX Runtime<\/a> with support for <a href=\"https:\/\/learn.microsoft.com\/en-us\/windows\/ai\/directml\/dml-intro\" target=\"_blank\" rel=\"noreferrer noopener\">Windows DirectML<\/a> along with cross-platform support across graphics processing unit (GPU), CPU, and even mobile hardware.<\/li>\n\n\n\n<li class=\"wp-block-list-item\">It is also available as an <a href=\"https:\/\/www.nvidia.com\/en-us\/ai\/\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA NIM<\/a> microservice with a standard API interface that can be deployed anywhere. And has been <a href=\"https:\/\/blogs.nvidia.com\/blog\/microsoft-open-phi-3-mini-language-models\" target=\"_blank\" rel=\"noreferrer noopener\">optimized for NVIDIA GPUs<\/a>.&nbsp;<\/li>\n<\/ul>\n\n\n\n<p class=\"wp-block-paragraph\">In the coming weeks, additional models will be added to Phi-3 family to offer customers even more flexibility across the quality-cost curve. <strong>Phi-3-small<\/strong> (7B) and <strong>Phi-3-medium<\/strong> (14B) will be available in the Azure AI model catalog and other model gardens shortly.\u202f\u202f&nbsp;<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Microsoft continues to offer the best models across the quality-cost curve and today\u2019s Phi-3 release expands the selection of models with state-of-the-art small models.<\/p>\n\n\n<div class=\"wp-block-msxcm-cta-block\" data-moray data-bi-an=\"CTA Block\">\n\t<div class=\"card d-block mx-ng mx-md-0\">\n\t\t<div class=\"row no-gutters\">\n\n\t\t\t\t\t\t\t<div class=\"col-md-4\">\n\t\t\t\t\t<img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"575\" src=\"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/04\/Azure_Hero_Hexagon_Offwhite_FullGrad-social-4.16-1024x575.webp\" class=\"card-img img-object-cover\" alt=\"abstract image\" srcset=\"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/04\/Azure_Hero_Hexagon_Offwhite_FullGrad-social-4.16-1024x575.webp 1024w, https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/04\/Azure_Hero_Hexagon_Offwhite_FullGrad-social-4.16-300x169.webp 300w, https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/04\/Azure_Hero_Hexagon_Offwhite_FullGrad-social-4.16-768x431.webp 768w, https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/04\/Azure_Hero_Hexagon_Offwhite_FullGrad-social-4.16.webp 1200w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/>\t\t\t\t<\/div>\n\t\t\t\n\t\t\t<div class=\"d-flex col-md\">\n\t\t\t\t<div class=\"card-body align-self-center p-4 p-md-5\">\n\t\t\t\t\t\n\t\t\t\t\t<h2>Azure AI Studio<\/h2>\n\n\t\t\t\t\t<div class=\"mb-3\">\n\t\t\t\t\t\t<p>Phi-3-mini is now available<\/p>\n\t\t\t\t\t<\/div>\n\n\t\t\t\t\t\t\t\t\t\t\t<div class=\"link-group\">\n\t\t\t\t\t\t\t<a href=\"https:\/\/aka.ms\/phi3-azure-ai\" class=\"btn btn-link text-decoration-none p-0\" target=\"_blank\">\n\t\t\t\t\t\t\t\t<span>Explore the release<\/span>\n\t\t\t\t\t\t\t\t<span class=\"glyph-append glyph-append-chevron-right glyph-append-xsmall\"><\/span>\n\t\t\t\t\t\t\t<\/a>\n\t\t\t\t\t\t<\/div>\n\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t<\/div>\n\n\t\t\t\t\t<\/div>\n\t<\/div>\n<\/div>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"groundbreaking-performance-at-a-small-size\">Groundbreaking performance at a small size<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Phi-3 models significantly outperform language models of the same and larger sizes on key benchmarks (see benchmark numbers below, higher is better). Phi-3-mini does better than models twice its size, and Phi-3-small and Phi-3-medium outperform much larger models, including GPT-3.5T.&nbsp;&nbsp;<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">All reported numbers are produced with the same pipeline to ensure that the numbers are comparable. As a result, these numbers may differ from other published numbers due to slight differences in the evaluation methodology. More details on benchmarks are provided in our <a href=\"https:\/\/arxiv.org\/abs\/2404.14219\" target=\"_blank\" rel=\"noreferrer noopener\">technical paper<\/a>.&nbsp;<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><em>Note: Phi-3 models do not perform as well on factual knowledge benchmarks (such as TriviaQA) as the smaller model size results in less capacity to retain facts.<\/em>&nbsp;<\/p>\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" src=\"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/04\/Screenshot-2024-04-23-102615.webp\" alt=\"\" class=\"wp-image-33514 webp-format\" data-orig-src=\"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/04\/Screenshot-2024-04-23-102615.webp\"><\/figure>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"safety-first-model-design\">Safety-first model design<\/h2>\n\n\n<div class=\"wp-block-msxcm-kicker-container\">\n\t<div class=\" wp-block-msxcm-kicker-block wp-block-msxcm-kicker--align-right\" data-bi-an=\"Kicker Right\">\n\t\t<p class=\"wp-block-msxcm-kicker__title small text-neutral-400 text-uppercase\">\n\t\t\tResponsible ai principles\t\t<\/p>\n\t\t<a\n\t\t\tclass=\"wp-block-msxcm-kicker__cta btn btn-link p-0 text-decoration-none\"\n\t\t\thref=\"https:\/\/www.microsoft.com\/en-us\/ai\/principles-and-approach\/\"\n\t\t\ttarget=\"_blank\"\t\t>\n\t\t\t<span>Learn about our approach<\/span> <span class=\"glyph-append glyph-append-xsmall wp-block-msxcm-kicker__glyph glyph-append-go\"><\/span>\n\t\t<\/a>\n\t<\/div>\n<\/div>\n\n\n\n<p class=\"wp-block-paragraph\">Phi-3 models were developed in accordance with the <a href=\"https:\/\/query.prod.cms.rt.microsoft.com\/cms\/api\/am\/binary\/RE5cmFl?culture=en-us&amp;country=us\" target=\"_blank\" rel=\"noreferrer noopener\">Microsoft Responsible AI Standard<\/a>, which is a company-wide set of requirements based on the following six principles: accountability, transparency, fairness, reliability and safety, privacy and security, and inclusiveness. Phi-3 models underwent rigorous safety measurement and evaluation, red-teaming, sensitive use review, and adherence to security guidance&nbsp;to help ensure that these models are responsibly developed, tested, and deployed in alignment with Microsoft&#8217;s standards and best practices. &nbsp;<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Building on our prior work with Phi models (\u201c<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/publication\/textbooks-are-all-you-need\/\" target=\"_blank\" rel=\"noreferrer noopener\">Textbooks Are All You Need<\/a>\u201d), Phi-3 models are also trained using high-quality data. They were further improved with extensive safety post-training, including reinforcement learning from human feedback (RLHF), automated testing and evaluations across dozens of harm categories, and manual red-teaming. Our approach to safety training and evaluations are detailed in our <a href=\"https:\/\/arxiv.org\/abs\/2404.14219\" target=\"_blank\" rel=\"noreferrer noopener\">technical paper<\/a>, and we outline recommended uses and limitations in the model cards. See the <a href=\"https:\/\/aka.ms\/phi3-hf\" target=\"_blank\" rel=\"noreferrer noopener\">model card collection<\/a>.<strong>&nbsp;<\/strong><\/p>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"unlocking-new-capabilities\">Unlocking new capabilities<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Microsoft\u2019s experience shipping copilots and enabling customers to transform their businesses with generative AI using <a href=\"https:\/\/azure.microsoft.com\/en-us\/solutions\/ai\/?gad_source=1&amp;gclid=Cj0KCQjwiYOxBhC5ARIsAIvdH50Eo9gnc-Tg7g6-I9Dt4Mv_npO543E8kV_NfioCTk4YzhdZfGEc85AaAt5pEALw_wcB#overview\/?ef_id=_k_Cj0KCQjwiYOxBhC5ARIsAIvdH50Eo9gnc-Tg7g6-I9Dt4Mv_npO543E8kV_NfioCTk4YzhdZfGEc85AaAt5pEALw_wcB_k_&amp;OCID=AIDcmm5edswduu_SEM__k_Cj0KCQjwiYOxBhC5ARIsAIvdH50Eo9gnc-Tg7g6-I9Dt4Mv_npO543E8kV_NfioCTk4YzhdZfGEc85AaAt5pEALw_wcB_k_\" target=\"_blank\" rel=\"noreferrer noopener\">Azure AI<\/a> has highlighted the growing need for different-size models across the quality-cost curve for different tasks. Small language models, like Phi-3, are especially great for:&nbsp;<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li class=\"wp-block-list-item\">Resource constrained environments including on-device and offline inference scenarios.<\/li>\n\n\n\n<li class=\"wp-block-list-item\">Latency bound scenarios where fast response times are critical.<\/li>\n\n\n\n<li class=\"wp-block-list-item\">Cost constrained use cases, particularly those with simpler tasks.<\/li>\n<\/ul>\n\n\n\n<p class=\"wp-block-paragraph\">For more on small language models, see our <a href=\"https:\/\/aka.ms\/AApbqsj\" target=\"_blank\" rel=\"noreferrer noopener\">Microsoft Source Blog<\/a>.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Thanks to their smaller size, Phi-3 models can be used in compute-limited inference environments. Phi-3-mini, in particular, can be used on-device, especially when further optimized with ONNX Runtime for cross-platform availability. The smaller size of Phi-3 models also makes fine-tuning or customization easier and more affordable. In addition, their lower computational needs make them a lower cost option with much better latency. The longer context window enables taking in and reasoning over large text content\u2014documents, web pages, code, and more. Phi-3-mini demonstrates strong reasoning and logic capabilities, making it a good candidate for analytical tasks.&nbsp;<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Customers are already building solutions with Phi-3. One example where Phi-3 is already demonstrating value is in agriculture, where internet might not be readily accessible. Powerful small models like Phi-3 along with <a href=\"https:\/\/learn.microsoft.com\/en-us\/azure\/data-manager-for-agri\/concepts-llm-apis\" target=\"_blank\" rel=\"noreferrer noopener\">Microsoft copilot templates<\/a> are available to farmers at the point of need and provide the additional benefit of running at reduced cost, making AI technologies even more accessible.&nbsp;&nbsp;<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">ITC, a leading business conglomerate based in India, is leveraging Phi-3 as part of their continued collaboration with Microsoft on the copilot for Krishi Mitra, a farmer-facing app that reaches over a million farmers. <\/p>\n\n\n\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<p class=\"wp-block-paragraph\">\u201c<em>Our goal with the Krishi Mitra copilot is to improve efficiency while maintaining the accuracy of a large language model. We are excited to partner with Microsoft on using fine-tuned versions of Phi-3 to meet both our goals\u2014efficiency and accuracy!<\/em>\u201d &nbsp;&nbsp;<\/p>\n<cite>Saif Naik, Head of Technology, ITCMAARS<\/cite><\/blockquote>\n\n\n\n<p class=\"wp-block-paragraph\">Originating in Microsoft Research, Phi models have been broadly used, with Phi-2 downloaded over 2 million times. The Phi series of models have achieved remarkable performance with strategic data curation and innovative\u202fscaling. Starting with Phi-1, a model used for Python coding, to Phi-1.5, enhancing reasoning and understanding, and then to Phi-2, a 2.7 billion-parameter model outperforming those up to 25 times its size in language comprehension.<sup>1<\/sup> Each iteration has leveraged high-quality training data and knowledge transfer techniques to\u202fchallenge conventional\u202fscaling laws.&nbsp;<\/p>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"get-started-today\">Get started today<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">To experience Phi-3 for yourself, start with playing with the model on <a href=\"https:\/\/aka.ms\/try-phi3\" target=\"_blank\" rel=\"noreferrer noopener\">Azure AI Playground<\/a>. You can also find the model on the <a href=\"https:\/\/huggingface.co\/chat\/models\/microsoft\/Phi-3-mini-4k-instruct\" target=\"_blank\" rel=\"noreferrer noopener\">Hugging Chat playground<\/a>. Start building with and customizing Phi-3 for your scenarios using the\u202f<a href=\"https:\/\/aka.ms\/phi3-azure-ai\">Azure AI Studio<\/a>. Join us to learn more about Phi-3 during a special\u202f<a href=\"https:\/\/aka.ms\/AIShow\/Phi-3\" target=\"_blank\" rel=\"noreferrer noopener\">live stream<\/a> of the AI Show.\u202f&nbsp;<\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\" \/>\n\n\n\n<p class=\"wp-block-paragraph\"><sup>1<\/sup> <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/blog\/phi-2-the-surprising-power-of-small-language-models\/\" target=\"_blank\" rel=\"noreferrer noopener\">Microsoft Research Blog, Phi-2: The surprising power of small language models, December 12, 2023<\/a>.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>We are excited to introduce Phi-3, a family of small open models with the capability of 10 times larger models, developed by Microsoft and optimized for Azure AI.<\/p>\n","protected":false},"author":42,"featured_media":33433,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"ms_queue_id":[],"ep_exclude_from_search":false,"_classifai_error":"","_classifai_text_to_speech_error":"","_alt_title":"","footnotes":"","msx_community_cta_settings":[]},"categories":[1454],"tags":[2671,2735,2747,3165,3167],"audience":[3057,3055,3056],"content-type":[1465],"product":[1803,2758],"tech-community":[],"topic":[],"coauthors":[2821],"class_list":["post-33425","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai-machine-learning","tag-ai","tag-copilot","tag-generative-ai","tag-language-models","tag-small-language-models-slms","audience-data-professionals","audience-developers","audience-it-implementors","content-type-announcements","product-azure-ai","product-azure-ai-studio","review-flag-1-1680286581-825","review-flag-2-1680286581-601","review-flag-3-1680286581-173","review-flag-5-1680286581-950","review-flag-7-1680286581-146","review-flag-anywh-1680286580-635","review-flag-microsofts","review-flag-new-1680286579-546","review-flag-partn-1680286579-901"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v27.2 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>Introducing Phi-3: Redefining what&#039;s possible with SLMs | Microsoft Azure Blog<\/title>\n<meta name=\"description\" content=\"We are excited to introduce Phi-3, a family of small open models offering groundbreaking performance, developed by Microsoft. Learn more.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Introducing Phi-3: Redefining what&#039;s possible with SLMs | Microsoft Azure Blog\" \/>\n<meta property=\"og:description\" content=\"We are excited to introduce Phi-3, a family of small open models offering groundbreaking performance, developed by Microsoft. Learn more.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/\" \/>\n<meta property=\"og:site_name\" content=\"Microsoft Azure Blog\" \/>\n<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/microsoftazure\" \/>\n<meta property=\"article:published_time\" content=\"2024-04-23T19:30:00+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2025-01-31T20:07:38+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/04\/Azure_Blog_Isometric_Illustration-10_1260x708.jpg\" \/>\n\t<meta property=\"og:image:width\" content=\"1260\" \/>\n\t<meta property=\"og:image:height\" content=\"708\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/jpeg\" \/>\n<meta name=\"author\" content=\"Misha Bilenko\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:image\" content=\"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/04\/Azure_Blog_Isometric_Illustration-10_1260x708.jpg\" \/>\n<meta name=\"twitter:creator\" content=\"@azure\" \/>\n<meta name=\"twitter:site\" content=\"@azure\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Misha Bilenko\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"5 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/#article\",\"isPartOf\":{\"@id\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/\"},\"author\":[{\"@id\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/author\/misha-bilenko\/\",\"@type\":\"Person\",\"@name\":\"Misha Bilenko\"}],\"headline\":\"Introducing Phi-3: Redefining what&#8217;s possible with SLMs\",\"datePublished\":\"2024-04-23T19:30:00+00:00\",\"dateModified\":\"2025-01-31T20:07:38+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/\"},\"wordCount\":1093,\"commentCount\":0,\"publisher\":{\"@id\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/#organization\"},\"image\":{\"@id\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/04\/Azure_Blog_Isometric_Illustration-10_1260x708.jpg\",\"keywords\":[\"AI\",\"Copilot\",\"Generative AI\",\"Language models\",\"Small language models (SLMs)\"],\"articleSection\":[\"AI + machine learning\"],\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/\",\"url\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/\",\"name\":\"Introducing Phi-3: Redefining what's possible with SLMs | Microsoft Azure Blog\",\"isPartOf\":{\"@id\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/#primaryimage\"},\"image\":{\"@id\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/04\/Azure_Blog_Isometric_Illustration-10_1260x708.jpg\",\"datePublished\":\"2024-04-23T19:30:00+00:00\",\"dateModified\":\"2025-01-31T20:07:38+00:00\",\"description\":\"We are excited to introduce Phi-3, a family of small open models offering groundbreaking performance, developed by Microsoft. Learn more.\",\"breadcrumb\":{\"@id\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/#primaryimage\",\"url\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/04\/Azure_Blog_Isometric_Illustration-10_1260x708.jpg\",\"contentUrl\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/04\/Azure_Blog_Isometric_Illustration-10_1260x708.jpg\",\"width\":1260,\"height\":708,\"caption\":\"Abstract illustration of a computer monitor\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Blog home\",\"item\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"AI + machine learning\",\"item\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/category\/ai-machine-learning\/\"},{\"@type\":\"ListItem\",\"position\":3,\"name\":\"Introducing Phi-3: Redefining what&#8217;s possible with SLMs\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/#website\",\"url\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/\",\"name\":\"Microsoft Azure Blog\",\"description\":\"Get the latest Azure news, updates, and announcements from the Azure blog. From product updates to hot topics, hear from the Azure experts.\",\"publisher\":{\"@id\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/#organization\",\"name\":\"Microsoft Azure Blog\",\"url\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/#\/schema\/logo\/image\/\",\"url\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/06\/microsoft_logo.webp\",\"contentUrl\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/06\/microsoft_logo.webp\",\"width\":512,\"height\":512,\"caption\":\"Microsoft Azure Blog\"},\"image\":{\"@id\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/#\/schema\/logo\/image\/\"},\"sameAs\":[\"https:\/\/www.facebook.com\/microsoftazure\",\"https:\/\/x.com\/azure\",\"https:\/\/www.instagram.com\/microsoftdeveloper\/\",\"https:\/\/www.linkedin.com\/company\/16188386\",\"https:\/\/www.youtube.com\/user\/windowsazure\"]},{\"@type\":\"Person\",\"@id\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/#\/schema\/person\/b2603da1afac705823964361ce9072c0\",\"name\":\"Kristin Gallagher\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/secure.gravatar.com\/avatar\/295fa37b6bb2bbf59603c38b6ac7a7b4b86cd0f736387182fa9d0117f52cdf5e?s=96&d=mm&r=gb83eb8c5c3f8feea9763b473dabe8524\",\"url\":\"https:\/\/secure.gravatar.com\/avatar\/295fa37b6bb2bbf59603c38b6ac7a7b4b86cd0f736387182fa9d0117f52cdf5e?s=96&d=mm&r=g\",\"contentUrl\":\"https:\/\/secure.gravatar.com\/avatar\/295fa37b6bb2bbf59603c38b6ac7a7b4b86cd0f736387182fa9d0117f52cdf5e?s=96&d=mm&r=g\",\"caption\":\"Kristin Gallagher\"},\"url\":\"https:\/\/azure.microsoft.com\/en-us\/blog\/author\/kristingallagher\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Introducing Phi-3: Redefining what's possible with SLMs | Microsoft Azure Blog","description":"We are excited to introduce Phi-3, a family of small open models offering groundbreaking performance, developed by Microsoft. Learn more.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/","og_locale":"en_US","og_type":"article","og_title":"Introducing Phi-3: Redefining what's possible with SLMs | Microsoft Azure Blog","og_description":"We are excited to introduce Phi-3, a family of small open models offering groundbreaking performance, developed by Microsoft. Learn more.","og_url":"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/","og_site_name":"Microsoft Azure Blog","article_publisher":"https:\/\/www.facebook.com\/microsoftazure","article_published_time":"2024-04-23T19:30:00+00:00","article_modified_time":"2025-01-31T20:07:38+00:00","og_image":[{"width":1260,"height":708,"url":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/04\/Azure_Blog_Isometric_Illustration-10_1260x708.jpg","type":"image\/jpeg"}],"author":"Misha Bilenko","twitter_card":"summary_large_image","twitter_image":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/04\/Azure_Blog_Isometric_Illustration-10_1260x708.jpg","twitter_creator":"@azure","twitter_site":"@azure","twitter_misc":{"Written by":"Misha Bilenko","Est. reading time":"5 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/#article","isPartOf":{"@id":"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/"},"author":[{"@id":"https:\/\/azure.microsoft.com\/en-us\/blog\/author\/misha-bilenko\/","@type":"Person","@name":"Misha Bilenko"}],"headline":"Introducing Phi-3: Redefining what&#8217;s possible with SLMs","datePublished":"2024-04-23T19:30:00+00:00","dateModified":"2025-01-31T20:07:38+00:00","mainEntityOfPage":{"@id":"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/"},"wordCount":1093,"commentCount":0,"publisher":{"@id":"https:\/\/azure.microsoft.com\/en-us\/blog\/#organization"},"image":{"@id":"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/#primaryimage"},"thumbnailUrl":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/04\/Azure_Blog_Isometric_Illustration-10_1260x708.jpg","keywords":["AI","Copilot","Generative AI","Language models","Small language models (SLMs)"],"articleSection":["AI + machine learning"],"inLanguage":"en-US","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/#respond"]}]},{"@type":"WebPage","@id":"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/","url":"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/","name":"Introducing Phi-3: Redefining what's possible with SLMs | Microsoft Azure Blog","isPartOf":{"@id":"https:\/\/azure.microsoft.com\/en-us\/blog\/#website"},"primaryImageOfPage":{"@id":"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/#primaryimage"},"image":{"@id":"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/#primaryimage"},"thumbnailUrl":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/04\/Azure_Blog_Isometric_Illustration-10_1260x708.jpg","datePublished":"2024-04-23T19:30:00+00:00","dateModified":"2025-01-31T20:07:38+00:00","description":"We are excited to introduce Phi-3, a family of small open models offering groundbreaking performance, developed by Microsoft. Learn more.","breadcrumb":{"@id":"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/#primaryimage","url":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/04\/Azure_Blog_Isometric_Illustration-10_1260x708.jpg","contentUrl":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/04\/Azure_Blog_Isometric_Illustration-10_1260x708.jpg","width":1260,"height":708,"caption":"Abstract illustration of a computer monitor"},{"@type":"BreadcrumbList","@id":"https:\/\/azure.microsoft.com\/en-us\/blog\/introducing-phi-3-redefining-whats-possible-with-slms\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Blog home","item":"https:\/\/azure.microsoft.com\/en-us\/blog\/"},{"@type":"ListItem","position":2,"name":"AI + machine learning","item":"https:\/\/azure.microsoft.com\/en-us\/blog\/category\/ai-machine-learning\/"},{"@type":"ListItem","position":3,"name":"Introducing Phi-3: Redefining what&#8217;s possible with SLMs"}]},{"@type":"WebSite","@id":"https:\/\/azure.microsoft.com\/en-us\/blog\/#website","url":"https:\/\/azure.microsoft.com\/en-us\/blog\/","name":"Microsoft Azure Blog","description":"Get the latest Azure news, updates, and announcements from the Azure blog. From product updates to hot topics, hear from the Azure experts.","publisher":{"@id":"https:\/\/azure.microsoft.com\/en-us\/blog\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/azure.microsoft.com\/en-us\/blog\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/azure.microsoft.com\/en-us\/blog\/#organization","name":"Microsoft Azure Blog","url":"https:\/\/azure.microsoft.com\/en-us\/blog\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/azure.microsoft.com\/en-us\/blog\/#\/schema\/logo\/image\/","url":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/06\/microsoft_logo.webp","contentUrl":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-content\/uploads\/2024\/06\/microsoft_logo.webp","width":512,"height":512,"caption":"Microsoft Azure Blog"},"image":{"@id":"https:\/\/azure.microsoft.com\/en-us\/blog\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/www.facebook.com\/microsoftazure","https:\/\/x.com\/azure","https:\/\/www.instagram.com\/microsoftdeveloper\/","https:\/\/www.linkedin.com\/company\/16188386","https:\/\/www.youtube.com\/user\/windowsazure"]},{"@type":"Person","@id":"https:\/\/azure.microsoft.com\/en-us\/blog\/#\/schema\/person\/b2603da1afac705823964361ce9072c0","name":"Kristin Gallagher","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/secure.gravatar.com\/avatar\/295fa37b6bb2bbf59603c38b6ac7a7b4b86cd0f736387182fa9d0117f52cdf5e?s=96&d=mm&r=gb83eb8c5c3f8feea9763b473dabe8524","url":"https:\/\/secure.gravatar.com\/avatar\/295fa37b6bb2bbf59603c38b6ac7a7b4b86cd0f736387182fa9d0117f52cdf5e?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/295fa37b6bb2bbf59603c38b6ac7a7b4b86cd0f736387182fa9d0117f52cdf5e?s=96&d=mm&r=g","caption":"Kristin Gallagher"},"url":"https:\/\/azure.microsoft.com\/en-us\/blog\/author\/kristingallagher\/"}]}},"msxcm_display_generated_audio":false,"msxcm_animated_featured_image":null,"distributor_meta":false,"distributor_terms":false,"distributor_media":false,"distributor_original_site_name":"Microsoft Azure Blog","distributor_original_site_url":"https:\/\/azure.microsoft.com\/en-us\/blog","push-errors":false,"_links":{"self":[{"href":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-json\/wp\/v2\/posts\/33425","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-json\/wp\/v2\/users\/42"}],"replies":[{"embeddable":true,"href":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-json\/wp\/v2\/comments?post=33425"}],"version-history":[{"count":0,"href":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-json\/wp\/v2\/posts\/33425\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-json\/wp\/v2\/media\/33433"}],"wp:attachment":[{"href":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-json\/wp\/v2\/media?parent=33425"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-json\/wp\/v2\/categories?post=33425"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-json\/wp\/v2\/tags?post=33425"},{"taxonomy":"audience","embeddable":true,"href":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-json\/wp\/v2\/audience?post=33425"},{"taxonomy":"content-type","embeddable":true,"href":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-json\/wp\/v2\/content-type?post=33425"},{"taxonomy":"product","embeddable":true,"href":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-json\/wp\/v2\/product?post=33425"},{"taxonomy":"tech-community","embeddable":true,"href":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-json\/wp\/v2\/tech-community?post=33425"},{"taxonomy":"topic","embeddable":true,"href":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-json\/wp\/v2\/topic?post=33425"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/azure.microsoft.com\/en-us\/blog\/wp-json\/wp\/v2\/coauthors?post=33425"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}