Meta just launched LlamaCon, their first-ever conference. I’m at the Meta HQ in Menlo Park. The mission was clear: beat OpenAI by accelerating the adoption of their open-source Llama models.
They released two big things: a Meta AI chatbot app for consumers and a Llama API for developers. The chatbot app is integrated with Meta platforms so you can share AI chats and get personalized responses based on app activity, basically a direct response to OpenAI’s social network.
The Llama API allows you to build apps with Llama models with just one line of code and no need to go through third-party cloud providers. This makes AI development way more accessible and puts Meta directly in competition with OpenAI’s API services.
Meta CEO Mark Zuckerberg talked about collaboration in the open-source space and called out labs like DeepSeek and Alibaba’s Qwen as allies in the fight against closed AI providers. He said open models combined and customized by developers can beat proprietary systems in quality and flexibility.
Some developers were hoping for a new reasoning model but Meta’s strategy is to build an open AI ecosystem not to release the most advanced model.
Meta may also have a regulatory edge. Under the EU AI Act, companies that distribute free and open-source AI systems get special treatment, a reason behind Meta’s push for openness.
In short, LlamaCon means Meta is going to redefine the AI space by being open, beating OpenAI and empowering developers with tools.
