During her first Llamacon AI developer conference on Tuesday, Meta announced an API for her series of Llama models: Llama API.
Available by limited overview, the LLAMA API allows developers to explore and experiment with products powered by different models of LAMA, by meta. Associated with Meta SDKs, it allows developers to create services, tools and applications focused on LAMA. Meta did not immediately share the API prices with Techcrunch.
The deployment of the API intervenes while Meta seeks to maintain an advance in the space of a fiercely competitive open model. While the Lama models have accumulated more than a billion downloads to date, according to Meta, rivals such as Deepseek and Qwen of Alibaba threaten to upset Meta’s efforts to establish a large -scale ecosystem with Llama.
The LLAMA API offers tools to refine and assess the performance of LLAMA models, starting with LLAMA 3.3 8B. Customers can generate data, train, then use Meta’s evaluation suite in the Llama API to test the quality of their personalized model.
Meta said that it will not use data from API customers to train the company’s own models and that the models built using the Llama API can be transferred to another host.
For developers who are built in addition to the recently published Llama 4 models of Meta, API Llama offers operating options of the model via partnerships with Cerebras and Groq. These “early experimental” options are “available by request” to help developers prototyper their AI applications, said Meta.
“By simply selecting Cerebras or Groq model names in the API, developers can … Take advantage of a rationalized experience with any use followed in one place,” Meta wrote in a blog article provided to Techcrunch. “(W) e I can’t wait to develop partnerships with additional providers to bring even more options to build on Llama.”
Meta said he would broaden access to the Llama API “in the coming weeks and months”.
