IBM is excited to announce the addition Meta’s latest generation of open models, Llama 4, to watsonx.ai. Llama 4 Scout and Llama 4 Maverick, the first mixture of experts (MoE) models released by Meta, provide frontier multimodal performance, high speeds, low cost, and industry leading context length.

The release of Llama 4 initiates a new era for the Llama series, introducing both an exciting evolution of the Llama architecture and an innovative approach to integrating different types of data modalities—including text, image, video—much earlier in the process than conventionally trained models. Both new models support a wide variety of text-in, text-out and image-in, text-out use cases.

With the introduction of these latest offerings from Meta, IBM now supports a total of 13 Meta models in the expansive library of foundation models available in watsonx.ai. In keeping with IBM’s open, multi-model strategy for generative AI, we continue to provide our platform customers with the most performant open models on the market today.

Click here to continue to read the full story:  https://www.ibm.com/new/announcements/Meta-llama-4-maverick-and-llama-4-scout-now-available-in-watsonx-ai