Manuel Candales

[Manuel Candales @manuelcandales]

Meta won gold at the IMO!

Being part of this has been the most beautiful professional experience I've ever had. Traveling to China and being at the IMO site in person made the experience truly complete.

Huge thanks to @shuchaobi for everything, and to @PoShenLoh for organizing the IMO Alumni Reunion.

[Shuchao Bi @shuchaobi]

To understand whether we're making genuine progress on reasoning, we entered our AI models in five international STEM Olympiad competitions this year.

The results:
🏅 Asian Physics Olympiad (APhO): Perfect score on the theory exam — gold medal
🏅 International Physics Olympiad (IPhO): Perfect score on the theory exam — gold medal
🥇 International Mathematical Olympiad (IMO): Gold medal, top 4% of human participants
🥇 International Chemistry Olympiad (IChO): Gold-medal level performance
🥇 Romanian Masters of Mathematics (RMM): Gold-medal level performance

Three of these (APhO, IPhO, IMO) were live competitions and our solutions were submitted under real competition conditions and graded by the official judges using the same marking criteria applied to student contestants.

A few things about the approach:
• Models were internally trained versions from the Muse Spark family
• Zero tool use: no search, no code interpreter, no calculator
• Multi-agent orchestration with parallel reasoning

We are excited about where this reasoning capability goes next; frontier research level across scientific domains and personal superintelligence.

Super grateful to the organizing committees of APhO, IPhO, and IMO for supporting our live participation. We have deep respect for the contestants and organizers behind these competitions. 🙏

And proud of the MSL team that pulled this together!

[Manuel Candales @manuelcandales]

#ExecuTorch is officially out of this world! 😁

Two Meta Quest 3 headsets were launched aboard NASA’s 35th SpaceX Commercial Resupply Services mission and arrived at the International Space Station today. 🚀

As part of the European Space Agency’s Mixed Reality Facility, astronauts will use them to rehearse spacewalks before venturing outside the station.

And, like all Meta Quest 3 headsets, they run AI models on-device using ExecuTorch!

[Manuel Candales @manuelcandales]

94% on ProofBench is wonder-full

[Manuel Candales @manuelcandales]

We jumped from 39% to 57% on ProofBench!
17%, 39%, 57% ... I am starting to see a pattern. 😄
What's next? 79%? 🚀
@shuchaobi @arnal_charles @AKiezun @cursuviam @Yong18850571

[Vals AI @ValsAI]

Muse Spark 1.2 is the first model to crack 60% on Finance Agent v2, our benchmark that gives models the job of a financial analyst.

At $0.77/test it is 6.7x cheaper than the previous #1, Opus 5 ($5.12), at twice the speed.

[Alexandr Wang @alexandr_wang]

great progress

[AI at Meta @AIatMeta]

To understand whether we're making genuine progress on reasoning, we entered our AI models in five STEM Olympiad competitions.

The results:
🏅 Asian Physics Olympiad (APhO): Perfect score, theory exam
🏅 International Physics Olympiad (IPhO): Perfect score, theory exam
🥇 International Mathematical Olympiad (IMO): Gold medal
🥇 International Chemistry Olympiad (IChO): Gold-medal-level performance
🥇 Romanian Masters of Mathematics (RMM): Gold-medal-level performance

The types of problems in the Olympiad competitions are exceptionally hard, demanding deep chains of reasoning, creative insight, and flawless argumentation. To test pure reasoning capability, we disallowed all tool use, meaning no search, no coding, and no calculator.

We have deep admiration for the contestants and committees behind these competitions, and are grateful for their support in enabling our participation.

[Vals AI @ValsAI]

Muse Spark 1.2 just cracked the top 5 on the Vals Index, at just $0.69 per test. This is 3x cheaper than Kimi and 10x or more cheaper than Fable, Opus, and 5.6 Sol.

[AI at Meta @AIatMeta]

Muse Spark 1.2 is available today in Muse Code and in Meta Model API, with expanded global access and many new features on the horizon.

➡️ Learn more: https://t.co/wjJYpGzVSX

[Shuchao Bi @shuchaobi]

We improved ProofBench from 17% to 39%, thanks to the lean enthusiasts Vivien Cabannes, @arnal_charles, Adam Kiezun, @cursuviam, @manuelcandales, Remi Munos etc...

[Vals AI @ValsAI]

Meta just released Muse Spark 1.1 and is the new SOTA on MedScribe and TaxEval, taking the top spot from Fable 5 while being 10x cheaper and twice as fast. Meta currently holds the top 2 spots on TaxEval

It is also the new #1 on Harvey's Legal Agent Bench, dethroning Grok 4.5 less than 24 hours after it took the top spot.

[AI at Meta @AIatMeta]

We’re excited to introduce Muse Spark 1.1, a significant upgrade from the first Muse Spark model we released earlier this year.

Along with this release, we are launching a public preview of the new Meta Model API where developers can access Muse Spark 1.1.

The model is also available now in "Thinking" mode in the Meta AI app and on https://t.co/wHkMPH82ZH.

Learn more: https://t.co/zGcA3XaWpN

[PyTorch @PyTorch]

Big congrats to the ExecuTorch team. Their paper just won the Best Industry Paper Award at @MLSysConf 2026. 🏆

ExecuTorch is PyTorch's framework for running models on-device, in production today across phones, wearables, desktops, and embedded systems. Backends include Apple, Arm, Qualcomm, MediaTek, Samsung, NXP, Cadence, Intel, NVIDIA, and others.

Catch the team Tuesday morning in the Best Paper Session ( AM PT) and at the poster session Thursday evening.

Paper: https://t.co/QlnMdybCOw
Discord: https://t.co/KsiEOO5ujC
X: https://t.co/lkYgcC8y6n
Project: https://t.co/v6wCXYtJkz
Schedule: https://t.co/vgNGvcoLHO

[PyTorch @PyTorch]

#ExecuTorch addresses fragmented native deployment for #AI agents as a #PyTorch native platform. It enables voice models across CPU, GPU, and NPU on Android, iOS, Linux, macOS & Windows

🔗 https://t.co/NeQQyUniL4

[Mistral for Developers @MistralDevs]

🎙️Run Voxtral Realtime locally with ExecuTorch!

💻Thanks to the ExecuTorch team, you can now deploy and run Voxtral Realtime efficiently and fully offline on your MacBook.

[AI at Meta @AIatMeta]

We’re advancing on-device AI with ExecuTorch, now deployed across devices including Meta Quest 3, Ray-Ban Meta, Oakley Meta Vanguard and Meta Ray-Ban Display.

By eliminating conversion steps and supporting pre-deployment validation in PyTorch, ExecuTorch accelerates the path from research to production, ensuring consistent, efficient AI across a diverse hardware ecosystem.

Read the full technical deep dive: https://t.co/CFqCVSj9b8

[PyTorch @PyTorch]

🚀 Excited to announce the release of ExecuTorch 1.0! Now, you can deploy PyTorch models seamlessly to edge devices—mobile, embedded, or desktop—without conversion or rewriting. ExecuTorch supports a wide range of hardware backends and model types, making on-device AI more accessible than ever.

🔗 Read the full blog from the PyTorch Team at Meta here: https://t.co/8NuNvcNmLb

#ExecuTorch #PyTorch #OpenSourceAI #AI #PyTorchCon

[Engineering at Meta @Meta_Engineers]

Over the last year, ExecuTorch – an open-source solution for on-device inference on mobile and edge devices – has been operational across key Meta apps, incrementally improving on-device ML performance, privacy, and efficiency.

Explore the technical buildups and results in our latest blog post: https://t.co/gy981FPo6t