Meta launched a new application programming interface on Tuesday for its Llama artificial intelligence models.

The company aims to make it easier for businesses and developers to integrate its technology and compete with offerings from AI heavyweights like OpenAI, Google, and China’s DeepSeek.

The announcement was made at Meta’s inaugural AI developer conference, where Chief Product Officer Chris Cox introduced the Llama API as a tool that simplifies AI adoption: “You can now start using Llama with one line of code,” he said during the keynote address.

APIs are essential tools that allow developers to customize and embed AI capabilities directly into their applications. For some rivals, such as OpenAI, API access has become the main commercial driver. With Llama, Meta appears to be charting a different path—releasing much of its model family free for developers and betting on long-term gains through innovation, ecosystem control, and increased engagement on its platforms.

Though Meta did not reveal pricing details, the company confirmed the Llama API would be made available initially to select users in a limited preview, with a broader rollout expected in the coming weeks.

In addition to the API, Meta released a standalone AI assistant app and reaffirmed plans to test a paid version of its AI chatbot later this year.

Meta’s approach offers more control and flexibility compared to some competitors. “You have full agency over these custom models—you control them in a way that’s not possible with other offers,” said Manohar Paluri, Meta’s Vice President of AI. “Whatever model you customize is yours to take wherever you want, not locked on our servers.”

The move comes as competition in the AI space intensifies, particularly from low-cost alternatives like DeepSeek, which has also released partially open-source models. The emergence of such rivals has raised investor concerns about the sustainability and expense of AI development at the top U.S. firms.

At the event, Meta engineers shared new strategies that have reduced costs and improved the efficiency of the latest Llama version. CEO Mark Zuckerberg welcomed the growing competition, saying it would prevent AI innovation from being monopolized by a few players.

“If another model, like DeepSeek, is better at something, then now as developers you have the ability to take the best parts of the intelligence from the different models and produce exactly what you need,” Zuckerberg said. “That’s going to be very powerful.”