ADeLe profiles AI models across a set of core abilities and compares them to task requirements. Published in Nature, this framework enables accurate prediction of model performance on tasks they have not encountered before: msft.it/6015QLEuN

Deploying AI across cultures is more than translation. Atlas is an interactive playbook that can help with cross-cultural deployment of AI applications. msft.it/6015QINXc

Multilingual AI means making hard tradeoffs. Translate or fine-tune? One model or many? Vibhasha is an interactive playbook to help you build multicultural apps and make those decisions with confidence. msft.it/6016QI3gu

Not every language has the AI infrastructure others do. Paza is an interactive playbook to help you build robust speech models for low-resource languages with benchmarking tools to help you choose the right approach. microsoft.github.io/Paza/

AI should work for everyone, wherever they are, and whatever language they speak. Introducing three interactive playbooks by Microsoft Research. These guides are for organizations building AI across languages, cultures, and markets. Meet Paza, Vibhasha, and Atlas. msft.it/6014QxJnm

AsgardBench evaluates whether embodied agents can revise their plans based on visual observations as tasks unfold. By focusing on perception-driven planning, it exposes key limitations and guides improvements in agent reliability. msft.it/6015QQ4fZ

GroundedPlanBench evaluates whether VLMs can plan actions and determine where they should occur. V2GP can improve both planning and spatial grounding, leading to more reliable robot behavior. Learn more: msft.it/6016QQQTg

Are machines truly intelligent? In Episode 1 of “The Shape of Things to Come,” technologists Subutai Ahmad & Nicolò Fusi join Microsoft’s Doug Burger to compare how large language models work with how the human brain learns & what it means for AI’s future. msft.it/6015QsiOZ

MicroLED datacenter networking, LLM-driven GPU design, real-time multimodal AI, agent memory and learning from experience, plus new benchmarks for interactive planning and a new podcast about the future of machine intelligence. msft.it/6016Qs9io

With advances happening fast, what will the future look like? Microsoft Research Labs lead Doug Burger explores this question via the lens of machine intelligence, climate & more in the podcast “The Shape of Things to Come.” Full episodes coming soon. msft.it/6013QW7P5

As AI agents transition from simple chatbots to complex autonomous systems, finding and fixing their errors gets harder. AgentRx is an automated diagnostic framework that pinpoints critical failures and supports more transparent, resilient agentic systems: msft.it/6012QlyiM

With advances happening fast, what will the future look like? Microsoft research lead Doug Burger explores this question via the lens of machine intelligence, climate & more in the podcast series “The Shape of Things to Come.” Full episodes coming soon. msft.it/6018Qcv5i

You’re no longer just doing the work—you’re directing. In our latest episode, @sineadbovell.bsky.social explores how the future of work is taking shape, and what it means as AI systems start to feel less like tools and more like teammates. Watch all episodes of On Second Thought: msft.it/6015QcqMd

PlugMem transforms AI agents’ interaction histories into structured, reusable knowledge. It integrates with any agent, supports diverse tasks and memory types, and maximizes decision quality while significantly reducing memory token use: msft.it/6017Qc9vv

Multimodal reasoning with Phi-4-reasoning-vision, new work on scaling LLM inference, benchmarking AI agents in network operations, cinematic video generation, adaptive evaluation for LLMs, and using AI to improve individual and population health. www.linkedin.com/pulse/phi-4-...

The latest Microsoft Research Forum episode is now available on-demand. Explore new ARO, Dion2, Magentic Marketplace, OptiMind, Agent Lightning, and Healthbots. Register to watch: events.microsoft.com/flow/ms/rese...

Vision-language models improve multimodal systems, but can make them slower, costlier, and harder to deploy. Learn how Phi-4-reasoning-vision-15B, a compact and fast multimodal reasoning model, blends strengths of different methods while reducing their limits: msft.it/6014Q5X0u

CORPGEN enables AI agents to manage dozens of interdependent tasks simultaneously in simulated workplace environments. It maintains performance under heavy multitasking, delivering up to 3.5x higher completion rates than leading baselines. msft.it/6015QbHoH

Long-term glass data storage advances, new work on transferable reasoning, multi-turn AI safety testing, multilingual AI design, and evaluating how models actually think. msft.it/6015QksvJ

As synthetic media grows, verifying what’s real, and the origin of content, matters more than ever. Our latest report explores media integrity and authentication methods, their limits, and practical paths toward trustworthy provenance across images, audio, and video. msft.it/6012QnGgi

Project Silica introduces new techniques for encoding data in borosilicate glass, as described in the journal Nature. These advances lower media cost and simplify writing and reading systems while supporting 10,000-year data preservation: msft.it/6017QVklt

Coming March 3 at 9:00 AM PT, our first Microsoft Research Forum episode of the year. New ARO, Dion2, Magentic Marketplace, OptiMind, Agent Lightning, and Healthbots. Register to watch: www.microsoft.com/en-us/resear...

We are pleased to announce that Doug Burger, Technical Fellow and Corporate Vice President at Microsoft Research, has been elected to the National Academy of Engineering Class of 2026 for advancing cloud-scale computing and networking with field-programmable systems. msft.it/6011Quo5H

Speech recognition for low-resource languages, medical foundation models, space-based ML, and brain-computer interfaces: msft.it/6016QPDpp

This research looks at why Predictive Inverse Dynamics Models often outperform standard Behavior Cloning in imitation learning. By using simple predictions of what happens next, PIDMs reduce ambiguity and learn from far fewer demonstrations. Learn more: msft.it/6018QMkdO

Microsoft Research unveils Paza, a human-centered speech pipeline, and PazaBench, the first leaderboard for low-resource languages. It covers 39 African languages and 52 models and is tested with communities in real settings. msft.it/6016QMDHe

Give this a second thought: when language is a barrier to access, AI can help play translator. Here’s how we’re building technology that reflects culture and lived experience. Explore recent episodes of 'On Second Thought' here: msft.it/6019QKh0f

Microsoft researchers received the AAAI-26 Outstanding Paper Award for LLM2CLIP, a vision-language framework that uses large language models as “teachers” to help CLIP better understand long, complex captions and achieve state-of-the-art multimodal performance. msft.it/6017QHtL1

In a new paper published in Communications of ACM (CACM), Microsoft researchers explore how a “web of agents” could enable an open, decentralized agentic economy—avoiding walled gardens that limit innovation, concentrate power, and reduce user choice. Read more: msft.it/6043QHUD9

AI can help generate medical image reports, but today’s models struggle with varying reporting schemes. Learn how UniRG uses reinforcement learning to boost performance of medical vision-language models: msft.it/6018Q1QTn

AI models that bring language-driven control to the physical world, new work on billion-scale vector search, spatially consistent video generation, tools for analyzing ML-driven cloud systems, and a CVPR 2026 workshop on multimodal AI agents. msft.it/6018Q1Hn8

Madan Musuvathi has been named an ACM Fellow by the Association for Computing Machinery, recognizing his foundational work in concurrency verification and his impact on modern machine learning systems design. Congratulations! msft.it/6014Q8Wvk

The question isn’t whether AI or doctors will shape healthcare. It’s how AI and clinicians can work together to deliver better care. Watch Episode 1 of On Second Thought with @sineadbovell.bsky.social for the full conversation. msft.it/6014Q8mIq

Rho-alpha, which translates natural language commands into control signals for robotic systems doing bimanual manipulation tasks, aims to make physical systems more adaptable by using physical sensing modalities like touch and continuous learning from human feedback. msft.it/6018QBUCn

Argos improves multimodal RL by evaluating whether an agent’s reasoning aligns with what it observes over time. The approach reduces visual hallucinations and produces more reliable, data-efficient agents for real-world applications: msft.it/6010QBEgy

“We’re really diverse yet medicine has to operate off of averages.” Futurist @sineadbovell.bsky.social explores with Microsoft’s Jonathan Carlson how AI is helping to unlock a new era of personalized medicine. Watch episode 1, live now: www.youtube.com/watch?v=WKrG...

OptiMind is a small language model that converts business operation challenges, described naturally, into mathematical formulations that optimization software can solve. It reduces formulation time & errors & enables fast, privacy-preserving local use: msft.it/6017t7ISB

A look at what’s next in AI, plus new research in multimodal reasoning, long-horizon robotics, scalable self-supervised learning, GPU optimization with AI, and interpretable LLM reasoning. msft.it/6012tfazi

2025 saw groundbreaking innovations including AI-powered materials discovery tools, protein structure modeling, and multilingual AI for underserved communities. Dive into our Year in Review for a look at these and other transformative advances. msft.it/6011tUVUJ

From AI-driven material discovery and protein modeling to agentic systems, gaming, and multilingual AI for the global majority, 2025 turned research into real-world impact. Explore the breakthroughs in Microsoft Research’s Year in Review. msft.it/6013tUPR3

Holoportation™ technology, enabling real-time 3D telecommunications, has evolved from the lab to real-world use. After a decade of refinement and real-world deployment, it's been released via open source license to encourage wider use and development: msft.it/6010toiZQ

The Agentic AI Research and Innovation (AARI) initiative brings Microsoft Research and Academic Researchers together to advance safe, robust agentic systems from foundational science to real-world impact through open, global collaboration. msft.it/6045tWLfv

Microsoft Chief Scientific Officer Eric Horvitz on what stood out at NeurIPS 2025, what’s next, and why these moments matter for the future of AI research.

The Microsoft Research Asia StarTrack Scholars Program is a three-month experience designed to foster global collaboration and accelerate frontier research. Applications are due today, December 15. Details on research fields, programs, and application procedures are available at: msft.it/6042tm1oI

ICYMI: Microsoft Chief Product Officer of Responsible AI Sarah Bird talks NeurIPS, the 20th anniversary of WiML, and what's next in RAI.

By decoupling how agents work from how they’re trained, Agent Lightning turns each step an agent takes into data for reinforcement learning. This makes it easy for developers to improve agent performance with almost zero code changes: msft.it/6011tczN7

Promptions helps developers add dynamic, context-aware controls to chat interfaces so users can guide generative AI responses. It lets users shape outputs quickly without writing long instructions: msft.it/6010tiw2r

The latest Microsoft Research Forum episode is now available on-demand. Explore purposeful research and its real-world impact. www.microsoft.com/en-us/resear...

We’re pleased to share Francesca Parmigiani has been named an Optica Fellow—one of photonics’ most competitive and prestigious distinctions—recognizing her pioneering work in high-speed optical communications and optical signal processing. Congratulations Francesca! msft.it/6014tgRGK

Using AI-generated virtual populations, Microsoft researchers uncovered hidden cellular patterns that could reshape how we understand and treat cancer. msft.it/6012t9guC