Llama Stack

Quick Start | Documentation | Colab Notebook | Discord

✨🎉 Llama 4 Support 🎉✨

We released Version 0.2.0 with support for the Llama 4 herd of models released by Meta.

👋 Click here to see how to run Llama 4 models on Llama Stack

Note you need 8xH100 GPU-host to run these models

pip install -U llama_stack

MODEL="Llama-4-Scout-17B-16E-Instruct"
# get meta url from llama.com
llama model download --source meta --model-id $MODEL --meta-url <META_URL>

# start a llama stack server
INFERENCE_MODEL=meta-llama/$MODEL llama stack build --run --template meta-reference-gpu

# install client to interact with the server
pip install llama-stack-client

CLI

# Run a chat completion
llama-stack-client --endpoint http://localhost:8321 \
inference chat-completion \
--model-id meta-llama/$MODEL \
--message "write a haiku for meta's llama 4 models"

ChatCompletionResponse(
    completion_message=CompletionMessage(content="Whispers in code born\nLlama's gentle, wise heartbeat\nFuture's soft unfold", role='assistant', stop_reason='end_of_turn', tool_calls=[]),
    logprobs=None,
    metrics=[Metric(metric='prompt_tokens', value=21.0, unit=None), Metric(metric='completion_tokens', value=28.0, unit=None), Metric(metric='total_tokens', value=49.0, unit=None)]
)

Python SDK

from llama_stack_client import LlamaStackClient

client = LlamaStackClient(base_url=f"http://localhost:8321")

model_id = "meta-llama/Llama-4-Scout-17B-16E-Instruct"
prompt = "Write a haiku about coding"

print(f"User> {prompt}")
response = client.inference.chat_completion(
    model_id=model_id,
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": prompt},
    ],
)
print(f"Assistant> {response.completion_message.content}")

As more providers start supporting Llama 4, you can use them in Llama Stack as well. We are adding to the list. Stay tuned!

🚀 One-Line Installer 🚀

To try Llama Stack locally, run:

curl -LsSf https://github.com/meta-llama/llama-stack/raw/main/install.sh | sh

Overview

Llama Stack standardizes the core building blocks that simplify AI application development. It codifies best practices across the Llama ecosystem. More specifically, it provides

• Unified API layer for Inference, RAG, Agents, Tools, Safety, Evals, and Telemetry.
• Plugin architecture to support the rich ecosystem of different API implementations in various environments, including local development, on-premises, cloud, and mobile.
• Prepackaged verified distributions which offer a one-stop solution for developers to get started quickly and reliably in any environment.
• Multiple developer interfaces like CLI and SDKs for Python, Typescript, iOS, and Android.
• Standalone applications as examples for how to build production-grade AI applications with Llama Stack.

%%{
	init: {
		'theme': 'base',
		'themeVariables': {
			'fontSize':'100px'
		}
	}
}%%
graph TD
	%% === Classes to control layout ===
	classDef inv rankSpacing:0,diagramPadding:0,nodeSpacing:1000,padding:0,opacity:0,stroke-width:0

	%% === Classes for color-coded nodes (scaled stroke width & corners) ===
	classDef darkGray fill:#ddd,stroke:#999,stroke-width:0px,rx:40px,ry:40px
	classDef agentYellow fill:#FDF3B6,stroke:#999,stroke-width:5px,rx:40px,ry:40px
	classDef violetBlock fill:#EEE8F4,stroke:#999,stroke-width:5px,rx:40px,ry:40px
	classDef lightGreen fill:#D8EDC1,stroke:#999,stroke-width:5px,rx:40px,ry:40px
	classDef telemetryGray fill:#E7E7E7,stroke:#999,stroke-width:5px,rx:40px,ry:40px
	%% === Top layer ===
	top[Llama Stack Client SDKs, CLI, User Interfaces]:::darkGray
	top ~~~ M1
	%% dummy right subgraph for alighnment and balance
	subgraph R
	end
	top ~~~ R
	subgraph Middle[" "]
		subgraph M1[" "]
			Agents:::agentYellow
			PostTraining[Post Training]:::violetBlock
		end

		M1 ~~~ M2
		subgraph M2[" "]
			classDef lightBlue fill:#C7E4F7,stroke:#999,stroke-width:5px,rx:40px,ry:40px
			classDef lightRed fill:#F8C1B1,stroke:#999,stroke-width:5px,rx:40px,ry:40px
			classDef lightOrange fill:#F9D591,stroke:#999,stroke-width:5px,rx:40px,ry:40px
			VectorIO:::lightBlue
			Inference:::lightRed
			Evals:::lightOrange
			SyntheticData[Synthetic Data]:::violetBlock
		end

		M2 ~~~ M3
		subgraph M3[" "]
			Safety:::lightGreen
			BatchInference[Batch Inference]:::violetBlock
			BatchAgents[Batch Agents]:::agentYellow
		end

		M3 ~~~ M4
		subgraph M4[" "]
		end
		M4 ~~~ M5
		subgraph M5[" "]
		%% === Dashed border classes (scaled stroke/dash/corners) ===
			classDef resBlue fill:#C9DCEC,stroke:#999,stroke-width:5px,rx:40px,ry:40px,stroke-dasharray:50 50
			classDef resGray fill:#EEE,stroke:#999,stroke-width:5px,rx:40px,ry:40px,stroke-dasharray:50 50
			classDef resYellow fill:#F6E3B3,stroke:#999,stroke-width:5px,rx:40px,ry:40px,stroke-dasharray:50 50
			classDef resGreen fill:#D8EDC1,stroke:#999,stroke-width:5px,rx:40px,ry:40px,stroke-dasharray:50 50
			VectorDBs:::resBlue
			Models:::resGray
			Shields:::resGreen
			Datasets:::resYellow
		end
		M5 ~~~ M6
		subgraph M6[" "]
        _[" "]:::inv
		end
		M6 ~~~ M7
		subgraph M7[" "]
			Telemetry:::telemetryGray
		end
	end
	M7 ~~~ SP
	%% dummy left subgraphs for alighnment and balance
	top ~~~ L
	subgraph L
	end
	L ~~~ L2
	subgraph L2
	end
	SP[  Service Providers  ]:::darkGray
	class Top,M1,M2,M3,M4,M5,M7,R,L,L2 inv
	classDef hr height:1,width:1500,fill:#EEE,stroke:#999,stroke-width:10,stroke-dasharray:10 50
	class L,R,M6 hr
	classDef MiddleC fill:#eee,rx:40px,ry:40px
	class Middle MiddleC

Llama Stack Benefits

• Flexible Options: Developers can choose their preferred infrastructure without changing APIs and enjoy flexible deployment choices.
• Consistent Experience: With its unified APIs, Llama Stack makes it easier to build, test, and deploy AI applications with consistent application behavior.
• Robust Ecosystem: Llama Stack is already integrated with distribution partners (cloud providers, hardware vendors, and AI-focused companies) that offer tailored infrastructure, software, and services for deploying Llama models.

By reducing friction and complexity, Llama Stack empowers developers to focus on what they do best: building transformative generative AI applications.

API Providers

Here is a list of the various API providers and available distributions that can help developers get started easily with Llama Stack.

API Provider Builder	Environments	Agents	Inference	Memory	Safety	Telemetry	Post Training
Meta Reference	Single Node	✅	✅	✅	✅	✅
SambaNova	Hosted		✅		✅
Cerebras	Hosted		✅
Fireworks	Hosted	✅	✅	✅
AWS Bedrock	Hosted		✅		✅
Together	Hosted	✅	✅		✅
Groq	Hosted		✅
Ollama	Single Node		✅
TGI	Hosted and Single Node		✅
NVIDIA NIM	Hosted and Single Node		✅
Chroma	Single Node			✅
PG Vector	Single Node			✅
PyTorch ExecuTorch	On-device iOS	✅	✅
vLLM	Hosted and Single Node		✅
OpenAI	Hosted		✅
Anthropic	Hosted		✅
Gemini	Hosted		✅
watsonx	Hosted		✅
HuggingFace	Single Node						✅
TorchTune	Single Node						✅
NVIDIA NEMO	Hosted						✅

Distributions

A Llama Stack Distribution (or "distro") is a pre-configured bundle of provider implementations for each API component. Distributions make it easy to get started with a specific deployment scenario - you can begin with a local development setup (eg. ollama) and seamlessly transition to production (eg. Fireworks) without changing your application code. Here are some of the distributions we support:

Distribution	Llama Stack Docker	Start This Distribution
Meta Reference	llamastack/distribution-meta-reference-gpu	Guide
SambaNova	llamastack/distribution-sambanova	Guide
Cerebras	llamastack/distribution-cerebras	Guide
Ollama	llamastack/distribution-ollama	Guide
TGI	llamastack/distribution-tgi	Guide
Together	llamastack/distribution-together	Guide
Fireworks	llamastack/distribution-fireworks	Guide
vLLM	llamastack/distribution-remote-vllm	Guide

Documentation

Please checkout our Documentation page for more details.

• CLI references
  • llama (server-side) CLI Reference: Guide for using the llama CLI to work with Llama models (download, study prompts), and building/starting a Llama Stack distribution.
  • llama (client-side) CLI Reference: Guide for using the llama-stack-client CLI, which allows you to query information about the distribution.
• Getting Started
  • Quick guide to start a Llama Stack server.
  • Jupyter notebook to walk-through how to use simple text and vision inference llama_stack_client APIs
  • The complete Llama Stack lesson Colab notebook of the new Llama 3.2 course on Deeplearning.ai.
  • A Zero-to-Hero Guide that guide you through all the key components of llama stack with code samples.
• Contributing
  • Adding a new API Provider to walk-through how to add a new API provider.

Llama Stack Client SDKs

Language	Client SDK	Package
Python	llama-stack-client-python
Swift	llama-stack-client-swift
Typescript	llama-stack-client-typescript
Kotlin	llama-stack-client-kotlin

Check out our client SDKs for connecting to a Llama Stack server in your preferred language, you can choose from python, typescript, swift, and kotlin programming languages to quickly build your applications.

You can find more example scripts with client SDKs to talk with the Llama Stack server in our llama-stack-apps repo.