Meta Announces LlamaCon: Its Inaugural Generative AI Developer Conference

On February 18, 2025, Meta revealed plans to host LlamaCon, its first developer conference focused entirely on generative AI. Named after Meta’s Llama series of AI models, LlamaCon is slated for April 29. At this event, Meta aims to showcase the latest advancements in their open-source AI projects, providing developers with tools and insights to develop innovative applications and products.

Meta has promised more details will follow shortly. For those interested, Meta’s more general annual developer event, Meta Connect, is scheduled for September.

Since adopting an open-source strategy for AI development several years ago, Meta has been cultivating a broad ecosystem around its AI technologies. Although the exact number of applications built on its platform isn’t public, Meta has mentioned that giants like Goldman Sachs, Nomura Holdings, AT&T, DoorDash, and Accenture leverage Llama. The model boasts hundreds of millions of downloads, with at least 25 partners, including Nvidia, Databricks, Groq, Dell, and Snowflake, enhancing the model’s capabilities, such as integrating with proprietary data or reducing latency.

However, Meta faced an unexpected challenge from DeepSeek, a Chinese AI firm that has introduced competitive open AI models. Reports suggest that Meta was taken aback by DeepSeek’s progress, particularly with a new model that might surpass the upcoming version of Llama. In response, Meta has reportedly organized emergency strategy sessions to understand and implement DeepSeek’s cost-effective model deployment techniques.

Meta has also committed to a hefty investment of up to $80 billion this year in AI, covering new hires, and building AI-specific data centers.

In other news, Mark Zuckerberg, Meta’s CEO, has outlined plans to roll out multiple Llama models in the coming months, ranging from reasoning models akin to OpenAI’s o3-mini, to multimodal models. He has also teased the potential for ‘agentic’ capabilities, where future models might execute tasks independently.

During Meta’s Q4 2024 earnings call, Zuckerberg expressed optimism about Llama’s potential, stating, “I think this very well could be the year when Llama and open source become the most advanced and widely used AI models… Our goal for Llama this year is to lead.”