Meta introduced a dedicated mobile app for its AI assistant featuring advanced reasoning, multilingual support and personalised interactions drawn from users’ Facebook and Instagram profiles. The app also offers a “discover” feed showcasing community queries and a voice-mode interface designed to foster more natural dialogue. Meta plans to integrate the assistant with its AI glasses and companion software later this year and will trial a paid subscription tier for premium features in the coming quarter.
Building on its open-source ethos, Meta will continue to provide core AI tools free of charge, while monetising select enhancements. This approach echoes past strategies but intensifies the platform’s direct competition with other consumer-facing AI chatbots.
Llama API opens to developers.
To attract software creators, Meta previewed its Llama API, inviting partners to join a limited-preview waitlist. Chris Cox, Meta’s Chief Product Officer, highlighted how the API will offer cloud-based access to the Llama model family, allowing companies to integrate advanced language and multimodal capabilities into bespoke applications. Initial partners will help shape pricing and performance benchmarks before a wider rollout later in 2025.
This marks a significant shift from the model-download strategy that has driven over 1.2 billion Llama downloads to date. By offering it as a managed service, Meta aims to lower technical barriers and capture developer mindshare in a marketplace dominated by a handful of providers.
New model advances and community grants
Attendees also learned about the expansion of the Llama 4 family, which now includes specialised variants utilising a mixture-of-experts architecture to optimise resource use for diverse tasks. These models support text, image and audio inputs, enabling more personalised multimodal experiences. Meta’s research division is further supporting innovation through grants totalling $1.5 million awarded to universities, startups and established companies working on transformative AI projects.
By seeding research and development efforts across academia and industry, Meta hopes to cultivate use cases that extend beyond social media and into fields such as healthcare, climate science and education.
Industry dialogue underscores AI’s wider impact
A highlight of LlamaCon was a fireside chat between Meta’s Mark Zuckerberg and Microsoft’s Satya Nadella. The two executives discussed AI’s potential to boost productivity and economic growth on a scale comparable to past industrial revolutions. Nadella revealed that up to 30% of code in some Microsoft repositories is now generated by AI, illustrating the deepening integration of these technologies in enterprise workflows.
With LlamaCon serving as a launchpad, Meta enters a critical phase in which it must convert developer interest into robust adoption while ensuring consumer trust. While the company’s open source credentials and vast user base provide a solid foundation, questions remain over how revenue-generating services will fare in a competitive landscape and how regulatory scrutiny might evolve.
Summary
In summary, Meta’s LlamaCon announcements reflect a dual-track strategy: broadening consumer access through a dedicated AI app and fostering a developer ecosystem via managed APIs and research support. Whether these efforts will translate into sustained leadership in AI will depend on execution, partner adoption and the wider industry’s response.
