Vincent Wehren

Product Architect @microsoftai | agentic browsing in Copilot and Cowork | Dad of 3 | Secret 90's era Dutch rocker | Defender of the Oxford comma

[Vincent Wehren @vincentwehren]

“Every review round finds something”

[Jacob Andreou @jacobandreou]

Couldn’t be more excited for Code in the Copilot super app coming soon. Everyone builds!

[Satya Nadella @satyanadella]

Super to see GPT-5.6 with Work IQ come to Copilot Chat, Cowork, M365 apps, GitHub, and Foundry today.

From multi-step agentic work to analysis and content creation, it brings stronger reasoning and higher-quality outputs without sacrificing efficiency.

[Satya Nadella @satyanadella]

The future of the firm is a learning loop in which human capital and token capital compound.

With our new Frontier Co., our ambition is to help every enterprise build its own AI capability, and to help create a frontier ecosystem where every organization can turn its knowledge, workflows, and judgment into its own AI systems that continuously improve. https://t.co/mvYhkRFyqa

[Vincent Wehren @vincentwehren]

Copilot Cowork now GA! And super proud my team has added browser control to Cowork's growing list of capabilities!

[Charles Lamanna @clamanna]

Copilot Cowork is now generally available!

Over the last few months of preview in Frontier, we’ve seen you use Cowork to help with so many different tasks. We’ve also been listening closely to your feedback and with GA, we’re bringing you more improvements + new features across model choice, extensibility through plugins, browser automation, and cost management controls.

Here’s a quick demo showing the latest updates in action - inspired by some of my own day-to-day tasks. You can also read more in the blog linked below...

[Charles Lamanna @clamanna]

Copilot Cowork is now generally available!

Over the last few months of preview in Frontier, we’ve seen you use Cowork to help with so many different tasks. We’ve also been listening closely to your feedback and with GA, we’re bringing you more improvements + new features across model choice, extensibility through plugins, browser automation, and cost management controls.

Here’s a quick demo showing the latest updates in action - inspired by some of my own day-to-day tasks. You can also read more in the blog linked below...

[Vincent Wehren @vincentwehren]

.Yep

[Microsoft AI @MicrosoftAI]

MAI-Code-1-Flash is now rolled out to 100% of GitHub Copilot Free, Education, Pro, Pro+, and Max subscribers in VS Code. Copilot CLI roll out and Enterprise/Business preview on its way. Give it a try and let us know what you think!

[Vincent Wehren @vincentwehren]

MAI Superintelligence team cookin’ 🔥

[Mustafa Suleyman @mustafasuleyman]

There are no shortcuts to the frontier. Disciplined, patient, meticulous attention to detail is critical. To give everyone a good sense of our progress we've published a very detailed technical report (109 pages!) outlining how we trained MAI-Thinking-1 and what we learned along the way. https://t.co/aCYyNybmUv

It's exceptionally strong in reasoning and SWE benchmarks. I’m super proud that it’s at 53% on SWE-Bench Pro, placing it right alongside Opus 4.6 on one of the toughest coding benchmarks. We have a lot more work to do as we get it into production and climb on more real world use cases.

We also launched six other world class models. Here's the line up: https://t.co/sYdMYnGSFD

- MAI-Transcribe-1.5 is the best transcription model in the world, with SOTA accuracy across 43 languages, beating out Gemini and OpenAI’s flagship transcription models. It offers the best accuracy, 5x faster than rival models. It's now in Microsoft Foundry, where it’s the fastest, most efficient and most cost‑effective transcription model of any hyper-scaler.

- MAI‑Voice‑2 our latest speech generation model, with native‑sounding delivery and fine‑grained emotional control, available in 15 languages with lots more coming soon. And MAI-Voice‑2‑Flash provides the best value and speed for ultra latency‑sensitive Voice Agents.

- MAI-Code-1-Flash is our new inference efficient coding model, especially tuned for VS Code and GitHub Copilot CLI. It's a brilliant, fast model at just 5B actives, and delivers 51% on SWE-Bench Pro. Super excited to get it into prod and climb on more real-world tasks.

- MAI-Image-2.5 and Flash are two super strong models that deliver a step change in quality and surpass Nano Banana 2 on the image editing leaderboard. MAI-Image-2.5 delivers super strong performance on H100s, enabling deployment on existing infrastructure, with flexibility to scale across GB200/GB300 systems.

We're hiring! We're a lean, fast-moving lab made of some of the world's most talented minds. Come join us as we work on our next generation of models!

[Microsoft AI @MicrosoftAI]

Seven new models launching at Build: let’s go!
Reasoning. Code. Image. Transcribe. Voice.

Built from scratch on a clean data lineage, designed for efficiency, working seamlessly as a family of models

Thread 🧵
#MSBuild

[Mustafa Suleyman @mustafasuleyman]

Super excited to announce seven new world-class MAI models today. They represent what we consider a new era in AI designed to keep you in control and on the frontier.
First is our text foundation model, MAI-Thinking-1, exceptionally strong on reasoning and SWE tasks.
- It’s a 35B active parameter MoE with a 256K context window. Independent human raters on Surge prefer it for overall quality in blind side-by-sides versus Sonnet 4.6, and it’s achieved 97% on AIME 2025, the key measure of its general-purpose reasoning abilities.
- It's at 53% on SWE Bench Pro, placing it right alongside Opus 4.6 on one of the toughest coding benchmarks.
- And since we co-designed our models with our own silicon, MAI-Thinking-1 is optimized on our MAIA 200 chip. Benchmarking head-to-head against the GB200, we see 30% better performance per dollar as well as a 1.4x performance-per-watt gain when running our MAI models on the MAIA 200 end-to-end.

Next is MAI-Image-2.5 and its Flash variant. Two super strong models now at #2 on the leaderboards, surpassing the score of Nano Banana 2 on image editing.

Last for now is MAI-Code-1-Flash, our new inference efficient coding model, especially tuned for VS Code and GitHub Copilot CLI.
- Code-1-Flash achieves 51% on SWE Bench Pro, despite having just 5B parameters, putting it closer to Haiku in size but cheaper in cost.

All of this is the foundation for Microsoft Frontier Tuning. It lets you customize our models to create custom, company-specific agents that only you control. You can make our model, your model. Your data. Your agents. Your moat.

Early adopters are already seeing a difference. When we tuned our models for McKinsey’s tasks, MAI delivered the highest win rate, outperforming GPT-5.5 on quality, while being 10x lower on cost.

Also really excited to be collaborating with the amazing team at Mayo Clinic to jointly train a new frontier AI model for healthcare.

Our announcements today mark another milestone on the road to humanist superintelligence. You can learn more and about our other new models in our latest blog: https://t.co/v65eop5Ixq

[Vincent Wehren @vincentwehren]

@craigaloewen So cool….

[Jordi Ribas @JordiRib1]

As Satya shared in today’s keynote at Build, we just launched Microsoft Web IQ - our next generation search engine for AI agents. It is a new suite of AI-native grounding APIs designed to connect AI systems with fresh, reliable web data for enhanced real-world intelligence including, web pages, news, images, and videos.

Search engines like Bing were built for humans, yet the next era of search will be for AI agents. Some estimates indicate that agents will generate 1000 times more queries than all search engines for humans combined in a few years, so not surprisingly, this is an exciting space with many competing solutions already.

To build Web IQ, our search team leveraged more than 20 years of technical innovations in Bing and re-architected the stack to serve agent queries most efficiently. Web IQ leads across the three things that matter most in this space: quality, latency, and token efficiency - as you can see in the graphs of the Bing blog linked below.

As an example, you may remember Harrier, our best-in-class embedding model I posted about a couple months back, which is still #1 in the relevant industry leaderboard. Harrier plays an important role in this system by helping optimize semantic retrieval, which strengthens the grounding pipeline. It’s just one of the many innovations that helped us build Web IQ.

Web IQ is also designed to honor publisher preferences by default, with systems and policies respecting publisher controls and content access preferences across the web.

The APIs in Web IQ are already powering nearly all AI agents and chatbots in the industry today, including Copilot and ChatGPT, so Microsoft is currently leading this new and rapidly growing category of search grounding for agents. I’m proud to share this important work and to continue driving innovation that is shaping the future of search.

[Vincent Wehren @vincentwehren]

This right here ⬇️

[Charles Lamanna @clamanna]

Software is shifting from apps built for people to agents that can reason and act. That means we need new programmability for M365. Today, we announced new Work IQ APIs – generally available June 16. #MSBuild

[Vincent Wehren @vincentwehren]

@cgillum I know them well!

[Vincent Wehren @vincentwehren]

RT @Microsoft: Introducing Project Solara, a new platform built from the ground up to power agent-driven experiences, including two new con…

[Pavan Davuluri @pavandavuluri]

Something new is coming for developers. And no, it’s not a new OS version 😅. See you at Build next week!

[Jacob Andreou @jacobandreou]

This is a fantastic model... Incredible work by the MAI Superintelligence Team

[Vincent Wehren @vincentwehren]

@steipete Love cmux + copilot cli. Use oh-my-posh with nerd fonts.

[Vincent Wehren @vincentwehren]

@jacobandreou Stayin’ AI-live, stayin’ AI-live…