Meta is launching a new program to inspire startups to adopt its Llama AI model.

The program is aimed at the startup Llama, providing the company with a “direct support” to Meta’s Llama team and in some cases funding the funding. Any formed U.S. company that raises less than $10 million in funding, has at least one developer with employees, and is building the generated AI application that is eligible to apply before the May 30 deadline.

“Members may receive up to $6,000 for six months to help them offset the cost of building and enhancing their generated AI solutions,” Meta wrote in a blog post. “Our experts will work closely with them to launch and explore advanced use cases for llamas that can benefit their startups.”

The Llama Startup program was launched as Meta tried to consolidate its lead in the competitive open model space. The tech giant's Llama model has received over a billion downloads to date, but competitors such as DeepSeek, Google and Alibaba's Qwen threaten to disrupt Meta's efforts to build a far-reaching model ecosystem.

The llamas have been helpless in the past few months and have suffered several setbacks.

The Wall Street Journal reported last week that Meta-element delayed the launch of flagship AI model Llama 4 Beymoth, a focus on key benchmarks for underperforming keys. In April, Meta had to boycott allegations of cheating on the popular group AI benchmark LM Arena. The company used a version of its Llama 4 Maverick model "optimized for dialogue" to score high scores in the LM arena, but publicly released different versions of Maverick.

Meta has huge ambitions for Llama – and its broader generative AI portfolio. Last year, the company predicted that the AI ​​products it generates will be between $2 billion and $3 billion in revenue in 2025 and between $460 billion and $1.4 trillion by 2035.

Meta has reached revenue sharing agreements with some companies that own the Llama model. The company recently launched an API for customizing Llama distributions. CEO Mark Zuckerberg said the company's AI AI assistant, Meta AI, may eventually display ads and provide additional features.

These products are expensive to build. In 2024, Meta's "Genai" budget exceeds $900 million, and this year's budget may exceed $1 billion. This does not include the infrastructure required to run and train the model. Meta had previously said it plans to spend $60 billion to $80 billion on capital expenditures in 2025, mainly on new data centers.