Meta unveiled an API for its artificial ielligence models called LLAMA API at the first conference of Llamacon artificial ielligence developers on Tuesday.
The LLAMA API is currely available in limited preview, but it allows developers to review and test products supported by various LLAMA models. It also allows developers to create LLAMA -based services, tools and applications.
The price of access to the LLAMA API is not yet specified
Meta has not yet announced the cost of access to this API. But the release of this API comes as Meta seeks to maiain its superiority in the competitive space of open source models. The LLAMA Meta models have had more than one billion downloads to date, but Meta has faced a fierce competition by Chinese companies such as Depsic and QWen Alibaba models, and these Chinese models can be a threat to the meta.
Meta also said that the LLAMA API will make the developers available for precise adjustme and performance evaluation of Lama models, which will start with the LLAMA 3.3 8B version. Customers can produce, educate, and then use the Meta API assessme set to test their custom model quality.
Llama API also provides Model Serving with the opportunity to develop with Cerebras and GREQ. According to Meta, these initial test options are available at the request of the user to help developers sampled their artificial ielligence programs.
In the end, Meta has promised developers that access to Llama API will expand over the coming mohs and weeks. Meta also unveiled the Meta AI dedicated app at the Llamacon conference.
