Meta is releasing a new tool that hopes will encourage developers to use their Llama model family for the next project. The company announced the Llama API at its inaugural Llamacon event in Menlo Park on Tuesday. The tool starts today as a limited free preview, providing developers with a place to try out Meta’s AI model, including the recently released Llama 4 Scout and Maverick Systems. It also makes it easy to create new API keys, which DEV can use for authentication purposes.
“We want to make it easier for you to start building with Llama, while also having full control over the model and weights without locking the API,” the company said in a blog post published during the event. To do so, the initial release of the Llama API includes tools developers can use to fine-tune and evaluate their applications.
Additionally, meta states that it does not use user prompts and model responses to train its own models. “When you’re ready, the models you build on the Camel API can be carried around, and we won’t let them lock on our servers wherever you want to host,” the company said. Meta hopes to roll out the tool to more users in the coming weeks and months.
Although Meta’s Llama model has been downloaded over a billion times, the company generally doesn’t see it as a leader in AI in the same way Openai and Anthropic. That doesn’t help push the company to capture the view of gaming LMARENA to make its Llama 4 model look better than it actually is.
