Meta enters AI race with Muse Spark, its major model since spending spree — here's what to know | Euronews

## Summary
By&nbsp Pascale Davies Published on 09/04/2026 - 12:35 GMT+2 Share Comments Share Facebook Twitter Flipboard Send Reddit Linkedin Messenger Telegram VK Bluesky Threads Whatsapp Meta has unveiled its first major AI model in nine months, following a $14.3 billion (€12.24 billion) investment spree and executive hiring push to rival OpenAI and Google. American tech company Meta has revealed its first major artificial intelligence (AI) model since it went on a spending spree nine months ago to boost its presence in the fiercely competitive AI market. Zuckerberg then embarked on a hiring spree, recruiting executives from rivals such as OpenAI, Anthropic, and Google. "Over the last nine months, Meta Superintelligence Labs rebuilt our AI stack from the ground up, moving faster than any development cycle we have run before," Meta wrote in a blog post on Wednesday. It is a powerful foundation, and the next generation is already in development." Muse Spark appears to be a major upgrade over Meta’s last big release, Llama 4, which came out in April 2025.

## Article Content
By&nbsp
Pascale Davies
Published on
09/04/2026 - 12:35 GMT+2
Share
Comments
Share
Facebook
Twitter
Flipboard
Send
Reddit
Linkedin
Messenger
Telegram
VK
Bluesky
Threads
Whatsapp
Meta has unveiled its first major AI model in nine months, following a $14.3 billion (€12.24 billion) investment spree and executive hiring push to rival OpenAI and Google.
American tech company Meta has revealed its first major artificial intelligence (AI) model since it went on a spending spree nine months ago to boost its presence in the fiercely competitive AI market.
ADVERTISEMENT
ADVERTISEMENT
Meta unveiled the model, called Muse Spark, on Wednesday, and claims it is smarter and faster than its earlier technologies.
The company, founded by Mark Zuckerberg, invested
$14.3 billion (€12.2 billion)
in the firm Scale AI in June 2025. It also hired its CEO and co-founder, Alexandr Wang, to oversee Meta Superintelligence Labs, which houses the company’s teams that work on foundational models.
Zuckerberg then embarked on a hiring spree, recruiting executives from rivals such as OpenAI, Anthropic, and Google.
"Over the last nine months, Meta Superintelligence Labs rebuilt our AI stack from the ground up, moving faster than any development cycle we have run before," Meta wrote in a
blog post
on Wednesday.
Related
China bans Manus founders from leaving country after Meta acquires AI startup and reviews deal
"This initial model is small and fast by design, yet capable enough to reason through complex questions in science, math, and health. It is a powerful foundation, and the next generation is already in development."
Muse Spark appears to be a major upgrade over Meta’s last big release, Llama 4, which came out in April 2025.
What do we know about the AI model?
Meta has said Muse Spark is capable of advanced reasoning capabilities and can answer complex questions, especially in science and maths. It added that the AI model is particularly good at providing medical advice.
"To improve Muse Spark's health reasoning capabilities, we collaborated with over 1,000 physicians to curate training data that enables more factual and comprehensive responses," the company's blog post read.
The new model will power the company’s digital assistant in the Meta AI app and website. As well as coming soon to Facebook, Instagram, WhatsApp and Messenger, it will also debut on the Ray-Ban Meta AI glasses.
Related
Mark Zuckerberg is building an AI bot to help run Meta and remain competitive in AI
The Meta AI app and site will also gradually feature a so-called contemplating mode for the most complicated queries and tasks, according to the company.
The contemplating mode will use several AI agents to help "reason in parallel," helping it "compete with the extreme reasoning modes of frontier models such as Gemini Deep Think and GPT Pro," the Meta technical blog read.
Zuckerberg said in a social media
post
that Meta's goal is to build AI products that "don't just answer your questions but act as agents that do things for you."
AI agents are designed to take autonomous actions to assist humans and do not require a human to tell them what to do, as they gather data based on user preferences.
This differs from AI chatbots, which are designed with conversation with humans in mind and serve as more of a co-pilot to assist humans.
One other point of interest marking a possible shift for the company is that Meta originally made its AI models
open source
, which generally means the software’s source code is available to everyone in the public domain to use, modify, and distribute.
But Meta’s new model is not available for download, meaning the technology is not open source.
Muse Spark is only available in the United States for the moment, the company said.
Go to accessibility shortcuts
Share
Comments
Read more
Tech News
Meet the robot dog that could one day sniff out life on Mars
Tech News
Telegram hosts network sharing non-consensual content in Spain, Italy
Tech News
What happens when AI becomes too powerful? Anthropic is finding out
Artificial intelligence
Meta
Mark Zuckerberg

---

## Expert Analysis

### Merits
- It added that the AI model is particularly good at providing medical advice. "To improve Muse Spark's health reasoning capabilities, we collaborated with over 1,000 physicians to curate training data that enables more factual and comprehensive responses," the company's blog post read.

### Areas for Consideration
N/A

### Implications
- The new model will power the company’s digital assistant in the Meta AI app and website.
- As well as coming soon to Facebook, Instagram, WhatsApp and Messenger, it will also debut on the Ray-Ban Meta AI glasses.
- Related Mark Zuckerberg is building an AI bot to help run Meta and remain competitive in AI The Meta AI app and site will also gradually feature a so-called contemplating mode for the most complicated queries and tasks, according to the company.
- The contemplating mode will use several AI agents to help "reason in parallel," helping it "compete with the extreme reasoning modes of frontier models such as Gemini Deep Think and GPT Pro," the Meta technical blog read.

### Expert Commentary
This article covers meta, model, company topics. Notable strengths include discussion of meta. Readability: Flesch-Kincaid grade 0.0. Word count: 671.