Llama on watsonx

Integrate Llama's powerful LLMs with the watsonx portfolio and accelerate enterprise AI adoption
Try Llama models on watsonx.ai
A screenshot of a computer screen displaying a Watsonx interface with various tabs and text boxes

Build with Llama on watsonx

Meta Llama 4 Maverick and Llama 4 Scout now available in watsonx.ai®
Flexibility to choose and deploy anywhere 

Leverage an advanced AI infrastructure, then deploy across cloud, hybrid or on-premises environments.
Simplified governance and management

Enable governance, streamline the AI model lifecycle, and foster collaboration across teams.
Customized for business needs

Find tailored engagement strategies for your enterprise, with the latest models tested and available from day one.
One-stop AI development studio

Use IBM® watsonx.ai® to develop and deploy applications with application programming interfaces (APIs), tools and models—turning ideas into reality.

Llama models available on watsonx.ai

Try Llama models on watsonx.ai
Llama 4 Scout
A compact multimodal model optimized for efficiency, capable of handling text and images.
Llama 4 Maverick
A high-performance multimodal model that utilizes a mixture-of-experts architecture for optimized resource use.
Llama 3.3 70B
A 70-billion-parameter text-only model optimized for commercial hardware, used for text generation and coding tasks.
Llama 3.2 90B Vision
A 90-billion-parameter multimodal model capable of processing both images and text, designed for advanced AI applications such as augmented reality and visual search.
Llama 3.2 11B Vision
An 11-billion-parameter multimodal model that processes images and text, suitable for real-time applications and edge devices.
Llama Guard 3 11B Vision
An 11-billion-parameter vision model focused on safety and content moderation, capable of analyzing images and text for harmful content.
Llama 3.2 1B
A lightweight 1-billion-parameter text-only model optimized for multilingual dialogue and tasks such as summarization—suitable for mobile devices.
Llama 3.2 3B
A 3-billion-parameter text-only model designed for multilingual dialogue and summarization tasks, balancing performance and efficiency.
Llama 3 405B
A large 405-billion-parameter text-only model optimized for multilingual dialogue, offering high performance for complex tasks.

Our clients

Resources

Unlock Enterprise AI Potential with Meta and IBM experts
Watch this recorded session to learn how industry leaders are using Llama models on watsonx.ai to tackle key business challenges and drive innovation.
Build an AI personal trainer with Meta Llama 4 using watsonx.ai
In this step-by-step tutorial, build a fitness app with an AI coach to create a customized training plan using Llama 4 models on watsonx.ai.
Behind-the-scenes of launching Llama 4 on IBM on Day 0
Hear from IBM engineers and product managers how IBM achieved a Day 0 launch, followed by a demo of Llama 4 on watsonx.ai.
Use Llama 3.2-90b-vision-instruct for multimodal AI queries in Python with watsonx
Apply a Llama vision model on watsonx.ai to computer vision tasks such as image captioning and visual question answering.
Meta's Llama 3.3 70b Instruct model now available on watsonx.ai
The updated Llama model offers performance comparable to the popular Llama 3.1 405B model, at a significantly lower cost.
Take the next step

Start using Llama models inside watsonx.ai today.

 Start your free trial Book a live demo