Meta✴ has announced its first developer conference dedicated to generative artificial intelligence (AI). The event, dubbed LlamaCon after Meta✴’s Llama family of generative AI models, is scheduled for April 29. The company plans to showcase its latest advances in open-source AI models to help developers build “the best apps and software products,” TechCrunch reports.
Image source: Meta✴
Meta✴ has been taking an open approach to AI development for several years, aiming to create an ecosystem of apps and platforms. While the company has not disclosed the exact number of apps or services built on Llama, the company has previously reported that organizations such as Goldman Sachs, Nomura Holdings, AT&T, DoorDash, and Accenture are using the Llama model.
The company says its model has been downloaded hundreds of millions of times, and at least 25 partners, including Nvidia, Databricks, Groq, Dell, and Snowflake, host Llama. Some have developed additional tools that allow AI models to reference proprietary data or operate with lower latency, for example.
Despite its apparent success, Meta✴ was unprepared for the explosive success of Chinese company DeepSeek, which released an AI model (also open source) capable of competing with Meta✴’s developments. According to rumors, Meta✴ believes that one of the latest versions of DeepSeek’s models may even surpass the next version of Llama, which is expected to be released in the coming weeks. Therefore, the company allegedly even urgently began to analyze DeepSeek’s methods for reducing the cost of operating and deploying models in order to apply this knowledge to its Llama developments.
Earlier, Meta✴ CEO Mark Zuckerberg announced several Llama models would be launched in the coming months, including models with reasoning capabilities similar to OpenAI’s o3-mini, as well as models with built-in multimodality. He also touched on the topic of AI agents (like OpenAI’s) that would emerge in the future and autonomously perform certain actions. “I think this could very well be the year that Llama and open-source models become the most advanced and widely used AI models,” Zuckerberg said. “We want to make sure that Llama is the leader this year.”
Meanwhile, Meta✴ faces legal and regulatory challenges. The company is embroiled in a lawsuit alleging that it used copyrighted material from books to train AI models without permission. Additionally, several EU countries have forced Meta✴ to delay or cancel plans to launch models in the region due to concerns about data privacy, creating additional obstacles for the company’s Llama ambitions overall.
Further details about LlamaCon, the company’s first event focused entirely on generative AI, will be announced soon, Meta✴ said. The company’s annual developer conference Meta✴ Connect will continue to take place in September.