Meta's new flagship model targets better coding and agentic performance
What's the story
Meta has launched Muse Spark 1.3, an upgraded version of its most advanced artificial intelligence (AI) model. The new system is available to developers via the Meta Model API, at the same rates as its predecessor, Muse Spark 1.2. The tech giant also plans to integrate this new system into its products such as Facebook and Instagram but has not provided a timeline for the rollout yet.
Model improvement
Muse Spark 1.3 on par with Claude Fable 5.1
Alexandr Wang, Meta's Chief AI Officer, has called the release of Muse Spark 1.3 a major leap in model performance.
He highlighted that the biggest improvements are seen in coding and "agentic" abilities, which allow AI systems to perform task sequences instead of just responding to individual prompts.
Wang claims Muse Spark 1.3 is on par with Anthropic's Claude Fable 5.1 and outperforms OpenAI's GPT-5.6 Sol on coding tasks.
Model capabilities
It can handle multiple workflows at once
Meta claims Muse Spark 1.3 can do more with about 25% fewer tokens, making it potentially more efficient to run.
The new model can handle multiple workflows at once, remember information over longer, complex tasks, and follow detailed instructions more consistently.
Wang said Muse Spark 1.3 is better at recognizing its limitations and can ask users for clarification or seek approval before taking irreversible actions.
Strategic shift
Meta under pressure to prove AI investments yield returns
The launch of Muse Spark 1.3 comes as Meta is under pressure to prove its massive AI investments can yield significant returns.
CEO Mark Zuckerberg has overhauled the company's AI strategy, including the establishment of Meta Superintelligence Labs and hiring Wang to lead it.
Despite investor scrutiny over its spending, Meta continues to invest heavily in AI infrastructure and talent to take on OpenAI and Anthropic.