Microsoft Foundry

[{"_1":2,"_3":-5,"_4":-5},"loaderData",{"_5":6,"_7":8,"_9":10},"actionData","errors","root",{"_301":302,"_303":304,"_305":306,"_307":25},"catalog/layout",{"_66":67,"_68":69,"_70":71,"_11":72},"catalog/all-models/AllModelsRoute",{"_11":12},"dehydratedState",{"_13":14,"_15":16},"mutations",[],"queries",[17],{"_18":19,"_20":21,"_22":23},"state",{"_26":27,"_28":29,"_30":31,"_32":-5,"_33":34,"_35":34,"_36":34,"_37":-5,"_38":-5,"_39":25,"_40":41,"_42":43},"queryKey",[24,25],"queryHash","[\"catalogCollection\",false]","catalogCollection",false,"data",{"_44":45},"dataUpdateCount",1,"dataUpdatedAt",1793193327337,"error","errorUpdateCount",0,"errorUpdatedAt","fetchFailureCount","fetchFailureReason","fetchMeta","isInvalidated","status","success","fetchStatus","idle","publishers",[46],{"_47":48,"_49":48,"_50":51,"_52":53,"_54":55,"_56":57,"_58":59,"_60":61,"_62":63,"_64":-7,"_65":25},"publisherName","xAI","displayName","iconLight","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","iconDark","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","description","Develops the Grok language models focused on reasoning, learning, and advancing reliable, high-performance AI systems.","publisherUIKey","","publisherIds",[],"registries",[],"disabledModels",{},"publisherContent","preview","flights",{"_292":25,"_293":25,"_294":25,"_295":71,"_296":71,"_297":71,"_298":71,"_299":71,"_300":71},"navigationData",{"_82":83,"_84":85},"unauthRedesign",true,{"_13":73,"_15":74},[],[75],{"_18":76,"_20":77,"_22":78},{"_26":80,"_28":29,"_30":81,"_32":-5,"_33":34,"_35":34,"_36":34,"_37":-5,"_38":-5,"_39":25,"_40":41,"_42":43},[79],"[\"modelIconConfig\"]","modelIconConfig","\n\n \n \n \n \n \n \n\n \n\n Microsoft Foundry\n\n \n \n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n\n \n \n \n\n

\n\n \n \n\n \n\n

\n\n \n \n \n\n",1793193327425,"featuredPublishers",[257,258,259,260,261,262],"publisherData",[86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119],{"_47":252,"_121":253,"_54":254,"_141":255,"_64":256},{"_47":247,"_121":248,"_54":249,"_141":250,"_64":251,"_235":71,"_65":71},{"_47":242,"_121":243,"_54":244,"_141":245,"_64":246},{"_47":48,"_121":240,"_54":55,"_141":241},{"_47":236,"_121":237,"_54":238,"_141":239},{"_47":231,"_121":232,"_54":233,"_64":234,"_235":71},{"_47":227,"_121":228,"_54":229,"_141":230},{"_47":222,"_121":223,"_54":224,"_141":225,"_64":226},{"_47":217,"_121":218,"_54":219,"_141":220,"_64":221},{"_47":212,"_121":213,"_54":214,"_141":215,"_64":216},{"_47":207,"_121":208,"_54":209,"_141":210,"_64":211},{"_47":203,"_121":204,"_54":205,"_141":206},{"_47":199,"_121":200,"_54":201,"_141":202},{"_47":195,"_121":196,"_54":197,"_141":198},{"_47":191,"_121":192,"_54":193,"_141":194},{"_47":187,"_121":188,"_54":189,"_141":190},{"_47":183,"_121":184,"_54":185,"_141":186},{"_47":179,"_121":180,"_54":181,"_141":182},{"_47":175,"_121":176,"_54":177,"_141":178},{"_47":171,"_121":172,"_54":173,"_141":174},{"_47":167,"_121":168,"_54":169,"_141":170},{"_47":163,"_121":164,"_54":165,"_141":166},{"_47":159,"_121":160,"_54":161,"_141":162},{"_47":155,"_121":156,"_54":157,"_141":158},{"_47":151,"_121":152,"_54":153,"_141":154},{"_47":147,"_121":148,"_54":149,"_141":150},{"_47":143,"_121":144,"_54":145,"_141":146},{"_47":138,"_121":139,"_54":140,"_141":142},{"_47":134,"_121":135,"_54":136,"_64":137},{"_47":131,"_121":132,"_54":133,"_65":71},{"_47":128,"_121":129,"_54":130,"_65":71},{"_47":125,"_121":126,"_54":127,"_65":71},{"_47":123,"_121":124},{"_47":120,"_121":122},"Van Dijk Lab","publisherId","van dijk lab","Mims Harvard","mims harvard","UW Institute for Protein Design","uw institute for protein design","Leading research organization focused on accelerating breakthroughs in protein engineering and biomolecular science, leveraging advanced computational methods and AI-driven discovery to design novel proteins for medicine, sustainability, and technology.","Inception","inception","Pioneers diffusion-based large language models (dLLMs) that generate text in parallel for ultra-low latency and efficiency, delivering breakthrough performance for code, text, and multimodal tasks in enterprise environments.","Voyage AI","voyage ai","Specializes in high-precision embedding and reranking models that power fast, multilingual search and retrieval for RAG applications, optimizing accuracy, latency, and cost for enterprise-scale AI solutions.","Wanglab","wanglab","Medical AI research group behind MedSAM2, a foundation model unifying 2D and 3D segmentation for precision medicine.","## Overview\nWangLab’s MedSAM2 is a breakthrough medical image segmentation foundation model designed for precision medicine. Built on the Segment Anything Model (SAM) 2.1 architecture, it uniquely unifies 2D and 3D segmentation across volumetric scans and video data. By reducing manual annotation costs by over 85% while preserving high accuracy, MedSAM2 accelerates clinical and research workflows where precise segmentation is traditionally time-consuming.\n\n### Why WangLab on Azure\n\nDeploying MedSAM2 on Azure unlocks secure, compliant infrastructure for healthcare applications. Pair the model with GPU-accelerated inference, Responsible AI tooling, and seamless workflow integration to bring state-of-the-art segmentation to cardiology, oncology, and surgery at scale.","Stability AI","stability ai","Specializes in open-source generative AI models for image and text generation, including Stable Diffusion.","keywords","Stable Diffusion, generative AI, open-source image generation","Sight Machine","sight machine","Offers AI-driven manufacturing analytics to improve operational efficiency and product quality.","manufacturing analytics AI, operational efficiency, Sight Machine solutions","Cerence","cerence","Provides AI-powered voice and speech recognition solutions for automotive and mobility applications.","automotive AI, voice recognition, Cerence mobility solutions","Bayer","bayer","Leverages AI for advancements in healthcare and agriculture, focusing on research and development.","healthcare AI, agricultural AI, Bayer AI research","Rockwell","rockwell automation","Develops AI-driven industrial automation solutions to optimize manufacturing processes and operations.","industrial automation AI, manufacturing optimization, Rockwell AI solutions","Saifr","saifr","Offers AI tools for content compliance and risk management in regulated industries.","content compliance AI, risk management tools, Saifr AI","NTT DATA","ntt data","Provides AI solutions across various industries, focusing on digital transformation and innovation.","industry AI solutions, digital transformation, NTT DATA AI","Bria","bria","Develops generative AI models for visual content creation, including images and videos.","generative visual AI, image and video generation, Bria AI models","Paige AI","paige","Specializes in AI-powered pathology solutions to enhance cancer diagnosis and treatment planning.","AI pathology, cancer diagnosis AI, Paige AI solutions","SDAIA","sdaia","Saudi Data and AI Authority developing AI solutions to drive national digital transformation.","Saudi AI initiatives, SDAIA models, digital transformation AI","JAIS","core42","Provides multilingual language models tailored for Arabic and English, facilitating diverse linguistic applications.","multilingual models, Arabic language AI, JAIS AI models","Nixtla","nixtla","Focuses on time-series forecasting models, offering tools like TimeGEN-1 for predictive analytics.","time-series forecasting, TimeGEN-1, predictive analytics","Deci AI","deci ai","Specializes in optimizing deep learning models for faster inference and deployment efficiency.","model optimization, efficient inference, Deci AI","AI21 Labs","ai21 labs","Develops advanced language models like Jurassic-2 for complex text generation and understanding tasks.","Jurassic-2, AI21 Labs, advanced language models","Gretel","gretel","Provides synthetic data generation tools to enhance privacy and augment datasets for machine learning.","synthetic data, privacy-preserving AI, data augmentation","Snowflake","snowflake","Cloud-based data platform that integrates with AI models for enhanced data processing and analysis.","cloud data platform, AI integration, Snowflake AI","Databricks","databricks","Provides a unified data analytics platform integrating AI model development and deployment capabilities.","data analytics platform, AI integration, Databricks AI","Cohere","cohere","Offers language models optimized for retrieval-augmented generation and enterprise applications.","retrieval-augmented generation, Cohere Command, enterprise language models","## Overview\nCohere Inc. is a Canadian‑founded generative‑AI company launched in 2019 by former Google Brain researchers. Built around large‑language‑model families such as Command R/R+, Command A, and high‑precision Rerank APIs, Cohere focuses on enterprise and regulated‑industry use cases—finance, healthcare, manufacturing, energy, and the public sector—rather than consumer chatbots. “Attention Is All You Need” transformer paper—continues to steer Cohere’s mission to deliver secure, domain‑tuned LLMs for real‑world business workflows.\n\n### Why Cohere on Azure\n\nDeploy Command or Rerank side‑by‑side with Azure AI Search, store embeddings in Azure Vector DBs, and enjoy unified billing plus data‑residency controls.","NVIDIA","nvidia","Offers GPU-optimized models and tools for high-performance AI applications across various domains.","GPU-accelerated AI, NVIDIA models, high-performance computing","## Overview\nNVIDIA’s Nemotron family supplies open‑weight reasoning models tuned for agentic workflows. The lineup scales from Nemotron Nano (8 B) for edge devices to Nemotron Ultra (253 B)—currently the top open model on reasoning leaderboards—while keeping permissive licensing.\n\n### Why NVIDIA on Azure\n\nCombine Nemotron with Azure GPU SKUs, Triton inference, and ML ops tooling to build high‑throughput agents without commercial licensing hurdles.","Mistral","mistral ai","French AI startup offering efficient and cost-effective language models, including Mistral 7B and Mixtral.","Mistral AI, Mixtral, efficient language models","## Overview\nMistral AI is a Paris‑based startup founded in April 2023 by former DeepMind and Meta researchers. In June 2024, it secured a €600 million Series B round that pushed its valuation to €6 billion, making it Europe’s highest‑valued generative‑AI company. The firm develops high‑performance, open‑weight LLMs such as Mistral 7B, Mixtral 8×22B, and the 123‑billion‑parameter flagship Mistral Large, which debuted through a strategic partnership with Microsoft Azure. Its sparse Mixture‑of‑Experts architecture and permissive licensing drive efficient inference and rapid adoption across both open‑source and enterprise communities.\n\n### Why Mistral on Azure\n\nGet European‑licensed open weights with Azure’s Content Safety, deterministic scaling, and pay‑as‑you‑go billing, ideal for regulated industries wanting transparent models.","Meta","meta","Open-source models like Llama 2, built for versatile language tasks and research applications.","Llama 2, Meta AI, open-source language models","## Overview\nMeta AI—the research arm of Meta Platforms—created the Llama family of open‑weight large language models, first launched in February 2023. Today the lineup ranges from lightweight 7 B models to the 405 B‑parameter Llama 3.1, with the natively multimodal Llama 4 Scout and Maverick unveiled in April 2025, stretching context windows to 256 K tokens and rivaling proprietary front‑runners on reasoning benchmarks. An open, source‑available license has catalyzed a 10× surge in cloud usage through 2024, making Llama the most widely deployed open LLM for edge and enterprise workloads alike.\n\n### Why Meta on Azure\n\nRun open weights under your own subscription, integrate with Azure GPU fleets, and fine‑tune with proprietary data while staying inside your compliance boundary.","Hugging Face","hugging face","Hosts a vast repository of open-source models and tools for natural language processing tasks.","open-source NLP models, Transformers library, Hugging Face Hub","Black Forest Labs","black forest labs","Research-driven company behind Flux, a family of open-weight diffusion models optimized for speed, fidelity, and controllability.","## Overview\nBlack Forest Labs (BFL) is a research-driven AI company focused on generative image models. Their flagship Flux family delivers state-of-the-art open-weight diffusion models optimized for speed, fidelity, and controllability. Flux models are designed for both creative workflows and enterprise-grade deployments, balancing cutting-edge research with permissive licensing.\n\n### Why Black Forest Labs on Azure\n\nRunning Flux on Azure lets you pair open-weight diffusion with enterprise infrastructure: scalable GPU SKUs, serverless deployment, and integrated observability. You get the freedom of open licensing combined with Azure’s compliance, RAI safeguards, and workflow integration—helping teams move from experimentation to production image generation without friction.","overrideDescription","DeepSeek","deepseek","Develops cost-effective large language models optimized for performance and efficiency.","DeepSeek R1, efficient LLMs, cost-effective AI models","xai","Grok, truthful AI, generative AI","Microsoft","microsoft","Microsoft's AI Superintelligence team builds enterprise-class AI models spanning reasoning, coding, image, voice, and transcription.","Microsoft AI models, MAI models, MAI-Thinking-1, reasoning models, Microsoft Foundry, Frontier Tuning","## Overview\nThe latest **MAI models** are trained from scratch with zero distillation on enterprise-grade, commercially licensed data, providing a clean lineage that is deployable into production with confidence. The flagship **MAI-Thinking-1**, a 35B-parameter reasoning model with a 256K context window, showcases the ability to compete with frontier models on tough benchmarks like **SWE-Bench Pro** at a fraction of the size and cost. Alongside it sit fast, inference-efficient models for coding and a chart-topping image model that already ranks among the best for editing.\n\n### Why Microsoft Models on Foundry\n\nMAI models are coming to **Microsoft Foundry**, where they run with first-party managed compute, granular quota, and zero data egress. The real differentiator is **Microsoft Frontier Tuning**: using reinforcement-learning environments, you adapt MAI models to your own data and workflows within your compliance boundary, so the resulting model stays yours.","Anthropic","anthropic","Anthropic is an AI research company that creates safe, reliable, interpretable, and steerable AI systems.","Claude, Clause-sonnet","## Overview\nAnthropic is an AI safety and research company that creates reliable, interpretable, and steerable AI systems. Anthropic’s flagship product is Claude, a frontier AI model trusted by leading enterprises and millions of users worldwide for complex tasks including coding, agents, financial analysis, research, and office tasks. Claude delivers exceptional performance while maintaining high safety standards.\n\n### Why Claude on Azure\n\nAnthropic’s Claude models on Azure combine Claude’s advanced reasoning and safety with seamless integration through Azure ecosystem.","Azure OpenAI","openai","Microsoft-hosted OpenAI models offering enterprise-grade security and compliance.","Azure OpenAI, gpt-5.5, gpt-5.4-mini, gpt-image-2, GPT-5, enterprise AI","## Overview\nOpenAI aims to create \"safe and beneficial\" artificial general intelligence, defined as highly autonomous systems that outperform humans at most economically valuable work. The organization leads today's AI with some of the world's most advanced multimodal foundation models, including its frontier **gpt‑5.5** with a 1M‑token context window, and continues to compress state‑of‑the‑art performance into smaller, less expensive variants like **gpt‑5.4‑mini**. Newer speech models such as **gpt‑realtime‑1.5** handle low‑latency transcription and voice, while **gpt‑image‑2** brings native image generation and editing at up to 4K resolution.\n\n### Why OpenAI on Azure\n\nAll OpenAI endpoints inherit enterprise‑grade Content Safety, flexible serverless or provisioned‑throughput deployments, Azure billing, and private‑network inference so you can move from prototype to production in days, not months.",{"_263":252,"_121":253,"_54":285,"_264":286,"_265":287},{"_263":247,"_121":248,"_54":249,"_264":248,"_265":280},{"_263":242,"_121":243,"_54":273,"_264":274,"_265":275},{"_263":222,"_121":223,"_54":224,"_264":223},{"_263":271,"_121":218,"_54":219,"_264":272},{"_263":236,"_121":237,"_54":238,"_264":237,"_265":266},"name","icon","featuredModels",[267,268,269,270],"DeepSeek-V4-Pro","DeepSeek-V4-Flash","DeepSeek-V3.2","DeepSeek-V3.1","Mistral AI","mistral","Proprietary AI models developed by Microsoft, tailored for various enterprise applications and integrated within Azure services.","phi",[276,277,278,279],"MAI-Voice-2","MAI-Image-2.5","MAI-Image-2.5-Flash","MAI-Transcribe-1.5",[281,282,283,284],"claude-opus-4-8","claude-sonnet-4-6","claude-opus-4-7","claude-haiku-4-5","Microsoft-hosted OpenAI models, including gpt-5.1 and Codex, offering enterprise-grade security and compliance.","aoai",[288,289,290,291],"gpt-5.5","gpt-5.3-codex","gpt-image-2","gpt-5.4-mini","enable_agents_catalog","enable_tools_catalog","enable_solutions_catalog","unauth_redesign_group","unauth_dynamic_hot_models","unauth_redesign","unauth_models_redesign","unauth_models_more_filters","unauth_model_version_picker","buildNumber","AiFoundry-2026-Oct-23-2026-10-23.3","platform","unknown","region","usgovarizona","prefersDarkTheme"]