Llama

The open-source AI models you can fine-tune, distill and deploy anywhere. Choose from our collection of models: Llama 3.1, Llama 3.2, Llama 3.3.

Download models

Try Llama on Meta AI

See how Llama is the leading open source model family

Learn more

LlamaCon

Save the date for an exclusive event exploring the exciting possibilities and potential of Llama.

4.29.25

Latest models

Llama includes multilingual text-only models (1B, 3B), including quantized versions, text-image models (11B, 90B) and Llama 3.3 70B model offering similar performance to the Llama 3.1 405B model, allowing developers to achieve greater quality and performance on text-based applications at a fraction of the cost.

Start building

Multilingual

Llama 3.1

•

8B: Light-weight, ultra-fast model you can run anywhere.

•

405B: Flagship foundation model driving widest variety of use cases

Download models

Lightweight and Multimodal

Llama 3.2

•

1B and 3B: Light-weight, efficient models you can run everywhere on mobile and on edge devices.

•

11B and 90B: Multimodal models that are flexible and can reason on high resolution images.

Download models

Multilingual

Llama 3.3

•

70B: Experience leading performance and quality at a fraction of the cost with our latest release.

Download models

Download our flagship foundation 405B model

Download models

Do more with Llama 3.2

Develop highly performative and efficient applications from our latest release.

Learn more

On-Device

Multimodal

Llama Stack

On-device

Use our 1B or 3B models for on device applications such as summarizing a discussion from your phone or calling on-device tools like calendar.

Learn more

Multimodal

Use our 11B or 90B models for image use cases such as transforming an existing image into something new or getting more information from an image of your surroundings.

Learn more

Llama Stack

Seamlessly build agentic applications from a comprehensive toolchain.

Learn more

On-device

Use our 1B or 3B models for on device applications such as summarizing a discussion from your phone or calling on-device tools like calendar.

Learn more

Llama Stack: a streamlined developer experience

Build faster, deploy anywhere and get the most out of the latest Llama models on day 1.

Learn more

For developers

Best practices included

Optimized support for agentic tool calling, safety guardrails, inference and much more, significantly lowering development costs.

Develop in your preferred language

Choose from python, node, kotlin, and swift programming languages to quickly build
your applications.

Choose from python, node, kotlin, and swift programming languages to quickly build your applications.

Develop & deploy anywhere

With a common API, choose any distribution and deploy on-prem, locally hosted, or even on-device at the edge.

For partners & distributors

A standard API

Requires fewer model level changes across versions accelerating time to market for new models and lowering engineering investment.

Interoperability with the ecosystem

Leverage the fast moving Llama ecosystem by building on a common API and incorporate new components faster.

Support for agentic components

Llama Stack releases natively support tool calling, safety guardrails, retrieval augmented generation, an inference loop and other agentic functionality.

For developers

Best practices included

Optimized support for agentic tool calling, safety guardrails, inference and much more, significantly lowering development costs.

Develop in your preferred language

Choose from python, node, kotlin, and swift programming languages to quickly build your applications.

Develop & deploy anywhere

With a common API, choose any distribution and deploy on-prem, locally hosted, or even on-device at the edge.

For Partners & Distributors

A standard API

Requires fewer model level changes across versions accelerating time to market for new models and lowering engineering investment.

Interoperability with the ecosystem

Leverage the fast moving Llama ecosystem by building on a common API and incorporate new components faster.

Support for agentic components

Llama Stack releases natively support tool calling, safety guardrails, retrieval augmented generation, an inference loop and other agentic functionality.

Model evaluations

We evaluated performance on over 150 benchmark datasets that span a wide range of languages. For the vision LLMs, we evaluated performance on benchmarks for image understanding and visual reasoning. In addition, we performed extensive human evaluations that compare Llama with competing models in real-world scenarios.

Learn more

Instruction-tuned benchmarks

Llama 3.3 instruction-tuned benchmarks

Lightweight instruction-tuned benchmarks

Vision instruction-tuned benchmarks

Leading with open source

Llama models have been downloaded over 600 million times on Hugging Face alone, making Llama the leading open source model family. Our partner ecosystem is helping build on this momentum by offering services through our Llama Stack, so anyone can build fast with Llama. And with this release of Llama 3.2, even more use cases can be supported.

Learn more

600M

downloads on Hugging Face to date

10x

growth since 2023

Partners enabling Llama

ARM, MediaTek and Qualcomm now allow you to run our lightweight models on your mobile or on-edge devices for the most capable "local agentic systems”. Dell is also offering their distribution with Llama Stack to help developers integrate their tool capabilities more seamlessly.

Llama Stack represents a significant leap in simplifying and standardizing the application of AI within enterprises across various use cases. With Llama Stack integrated into Dell AI Factory, we're setting the stage for widespread adoption of open models on-premises.

“Llama Stack represents a significant leap in simplifying and standardizing the application of AI within enterprises across various use cases. With Llama Stack integrated into Dell AI Factory, we're setting the stage for widespread adoption of open models on-premises.”

Ihab Tarazi, CTO, Dell Technologies

Community stories

Learn how partners across the community are putting Llama to use in real life.

Learn more

Data privacy

AI Companion, a generative AI assistant leveraging Zoom's LLM built on Llama 2 and third-party models, enhances productivity and collaboration through chat, email and meeting summaries, with data privacy and AI control.

Zoom

Productivity

DoorDash uses Llama to streamline and accelerate daily tasks, such as leveraging its internal knowledge base to answer complex questions for the team and delivering actionable pull request reviews to improve its codebase.

DoorDash

Contextual Understanding

The creator of Pokémon GO launched their AR-first game Peridot, which uses Llama 2 to generate environment-specific reactions and animations based on what the pet characters are interacting with and seeing in the real world.

Niantic

Solving business needs

KPMG leveraged Llama for multiple use cases across industries. They helped a US bank's wholesale credit team explore secure open-source LLMs options to help enable faster and more efficient review of complex loan applications to help position them to take automation to the next level.

KPMG

Our partner ecosystem

Latest Llama updates

5 Steps to Getting Started with Llama 2 graphic

Stay up-to-date

Our latest updates delivered to your inbox

Subscribe to our newsletter to keep up with the latest Llama updates, releases and more.

Llama

Llama

The open-source AI models you can fine-tune, distill and deploy anywhere. Choose from our collection of models: Llama 3.1, Llama 3.2, Llama 3.3.

See how Llama is the leading open source model family

LlamaCon

Save the date for an exclusive event exploring the exciting possibilities and potential of Llama.

4.29.25

Latest models

Llama 3.1

Llama 3.2

Llama 3.3

Download our flagship foundation 405B model

Do more with Llama 3.2

Llama Stack: a streamlined developer experience

Build faster, deploy anywhere and get the most out of the latest Llama models on day 1.

For developers

Best practices included

Develop in your preferred language

Develop & deploy anywhere

For partners & distributors

A standard API

Interoperability with the ecosystem

Support for agentic components

For developers

Best practices included

Develop in your preferred language

Develop & deploy anywhere

For Partners & Distributors

A standard API

Interoperability with the ecosystem

Support for agentic components

Model evaluations

Instruction-tuned benchmarks

Leading with open source

600M

10x

Partners enabling Llama

Llama Stack represents a significant leap in simplifying and standardizing the application of AI within enterprises across various use cases. With Llama Stack integrated into Dell AI Factory, we're setting the stage for widespread adoption of open models on-premises.

“Llama Stack represents a significant leap in simplifying and standardizing the application of AI within enterprises across various use cases. With Llama Stack integrated into Dell AI Factory, we're setting the stage for widespread adoption of open models on-premises.”

Community stories

Data privacy

Productivity

Contextual Understanding

Solving business needs

Our partner ecosystem

Introducing Multimodal Llama 3.2 with Amit Sangani

Llama developer video: Knowledge distillation with 405B

Meta Connect 2024 AI session videos

Llama 3.2 1B and 3B models now quantized

Our latest updates delivered to your inbox