Quick Start

Self-Hosting Documentation Access

This section requires a password to access. Interested in self-hosting? Contact sales to learn more.

Password:

Access Documentation

Self-Hosting Documentation Access Granted Logout

Note

This page assumes that you are deploying the latest version of LlamaCloud. Please refer to the release notes for more information about previous versions.

Before You Get Started

Welcome to LlamaCloud! Before you get started, please make sure you have the following prerequisites:

• LlamaCloud License Key. To obtain a LlamaCloud License Key, please contact us at support@llamaindex.ai.
• Kubernetes cluster >=1.28.0 and a working installation of kubectl.
  • We are largely aligned with the versions supported in EKS, AKS, and GKE.
  • For a local evaluation deployment, any conformant local cluster works (kind, k3d, minikube, Docker Desktop). See the note below.
• Helm v3.7.0+
  • To install Helm, please refer to the official Helm Documentation.
• OpenAI API Key or Azure OpenAI Credentials. Configuring OpenAI credentials is the easiest way to get started with your deployment.
  • LlamaCloud tries to meet you at your organization’s needs and supports configuring more than OpenAI LLMs in including Anthropic, Bedrock, Vertex AI, and more.
  • Please refer to the docs in the Configuration section of the sidebar to learn more about configuring other LLMs.
• File Storage: LlamaCloud stores files in object storage (S3, Azure Blob, GCS, or any S3-compatible store such as MinIO).
  • Please follow the File Storage documentation to configure your deployment, and create the eight buckets it lists before installing the chart.
  • Set bucket lifecycle rules as part of that setup. LlamaCloud does not delete parse or extract output itself, so without them a deployment retains every parsed document indefinitely. See Retention and Bucket Lifecycle.
• Authentication Settings:
  • OIDC. OIDC is our recommended authentication mode for production deployments.
  • Basic Auth (email/password): As of July 24th, 2025 (v0.5.0), we support both oidc and basic authentication methods. This is a simpler authentication mode more suitable for staging deployments.
  • For more information, please refer to the Authentication Modes documentation.
• Credentials to External Services (See below).

Tip

Evaluating on a single machine? The same chart runs on a local cluster with no cloud account. Two substitutions cover everything the list above assumes is cloud-hosted:

• Object storage: run MinIO in the cluster and point S3Proxy at it. See the MinIO / S3-compatible tab in File Storage.
• Authentication: use Basic Auth instead of standing up an OIDC provider. See Authentication Modes.

The backing stores install from the Bitnami charts exactly as documented in Database and Queues, and Temporal comes up as a subchart. Size the machine against the Hardware Requirements below. A default install needs ~14 allocatable vCPUs, which is the constraint most laptops hit first.

External Services

LLamaCloud requires the following external services to be available: Postgres, MongoDB, Temporal, RabbitMQ and Redis.

Tip

Temporal can be deployed as a subchart within LlamaCloud by setting temporal.deploy: true. This is the recommended approach for getting started quickly. For production deployments, you may prefer to manage Temporal separately.

Please follow the Database and Queues documentation to configure these services for your deployment.

Caution

Deploy these external services before installing the LlamaCloud chart. The minimal values.yaml below points at in-cluster hostnames (postgresql, mongodb, rabbitmq, redis-master) that must already resolve inside the release namespace.

If you run them in-cluster with the Bitnami charts, your cluster also needs a default StorageClass — or you must set global.storageClass on each store — otherwise their PersistentVolumeClaims stay Pending and the pods never start. Amazon EKS does not ship a default StorageClass out of the box.

Hardware Requirements

• Linux Instances running x86 CPUs
  • We currently only linux/amd64 images. arm64 is not supported at this moment.
• Ubuntu >=22.04 on the worker nodes
• >=14 allocatable vCPUs
• >=80 GiB allocatable memory

These are cluster-wide allocatable minimums for an install with default values. The chart’s default resource requests sum to 12.5 vCPU / 47.5 GiB on their own, so a 12-vCPU cluster cannot schedule a default install at all. The figures above add headroom for system pods and scheduling slack. Two things push the requirement higher:

• S3Proxy (config.storageBuckets.s3proxy.enabled: true) adds a 500m / 512Mi sidecar to every S3-consuming pod, roughly +3 vCPU across a default install.
• Containerized dependencies. Running PostgreSQL, MongoDB, RabbitMQ, Redis, or Temporal on the same cluster adds their requests on top (see the next section).

Warning: LlamaParse, LlamaIndex’s proprietary document parser, can be a very resource-intensive deployment to run, especially if you want to maximize performance.

Note

GPU is required for OCR. Set config.parseOcr.gpu: true and provide GPU nodes. Each OCR pod then requests nvidia.com/gpu: 1 and schedules only on those nodes.

Running OCR on CPU (config.parseOcr.gpu: false) is deprecated and will be removed in a future release. It is dramatically slower than the GPU backend and is not supported for production workloads. Plan GPU capacity for any deployment that parses scanned or image-based documents.

Configure and Install Your Deployment

This section will walk you through the steps to configure a minimal LlamaCloud deployment.

Minimal values.yaml configuration

To get a minimal LlamaCloud deployment up and running, you can create a values.yaml file with the following content:

OpenAI w/ OIDC Auth

  license:
    key: <LLAMACLOUD-LICENSE-KEY>

  postgresql:
    host: "postgresql"
    port: "5432"
    database: "llamacloud"
    username: <POSTGRES-USERNAME>
    password: <POSTGRES-PASSWORD>

  mongodb:
    host: "mongodb"
    port: "27017"
    username: <MONGODB-USERNAME>
    password: <MONGODB-PASSWORD>

  rabbitmq:
    scheme: "amqp"
    host: "rabbitmq"
    port: "5672"
    username: <RABBITMQ-USERNAME>
    password: <RABBITMQ-PASSWORD>

  redis:
    scheme: "redis"
    host: "redis-master"
    port: "6379"
    db: 0

  # Deploy Temporal as a subchart (host/port auto-configured)
  temporal:
    deploy: true

  # Temporal subchart configuration (uses same PostgreSQL instance).
  # The Temporal store password is taken from your postgresql.password above —
  # the chart generates the temporal-postgresql-secret automatically.
  temporal-subchart:
    server:
      config:
        persistence:
          defaultStore: default
          visibilityStore: visibility
          datastores:
            default:
              sql:
                connectAddr: "postgresql:5432"
                databaseName: temporal
                user: <POSTGRES-USERNAME>
            visibility:
              sql:
                connectAddr: "postgresql:5432"
                databaseName: temporal_visibility
                user: <POSTGRES-USERNAME>

  config:
    llms:
      openAi:
        apiKey: <OPENAI-APIKEY>

    frontend:
      enabled: true
    parseOcr:
      # Each OCR pod requests `nvidia.com/gpu: 1` and schedules only on GPU
      # nodes. Running OCR on CPU (gpu: false) is deprecated and will be
      # removed in a future release.
      gpu: true

    # Required. Without this the chart falls back to provider "aws" and the
    # literal default bucket names, which are almost certainly not yours.
    # See the File Storage guide for S3-compatible and non-AWS backends.
    storageBuckets:
      provider: "aws"
      parsedDocuments: <YOUR-BUCKET>
      parsedEtl: <YOUR-BUCKET>
      parsedExternalComponents: <YOUR-BUCKET>
      parsedFileParsing: <YOUR-BUCKET>
      parsedRawFile: <YOUR-BUCKET>
      parseOutput: <YOUR-BUCKET>
      parsedFileScreenshot: <YOUR-BUCKET>
      extractOutput: <YOUR-BUCKET>
      parseFileUpload: <YOUR-BUCKET>
      parseFileOutput: <YOUR-BUCKET>

    authentication:
      oidc:
        enabled: true
        discoveryUrl: "https://login.microsoftonline.com/<TENANT-ID>/v2.0/.well-known/openid-configuration"
        clientId: <CLIENT-ID>
        clientSecret: <CLIENT-SECRET>

Azure OpenAI w/ OIDC Auth

  license:
    key: <LLAMACLOUD-LICENSE-KEY>

  postgresql:
    host: "postgresql"
    port: "5432"
    database: "llamacloud"
    username: <POSTGRES-USERNAME>
    password: <POSTGRES-PASSWORD>

  mongodb:
    host: "mongodb"
    port: "27017"
    username: <MONGODB-USERNAME>
    password: <MONGODB-PASSWORD>

  rabbitmq:
    scheme: "amqp"
    host: "rabbitmq"
    port: "5672"
    username: <RABBITMQ-USERNAME>
    password: <RABBITMQ-PASSWORD>

  redis:
    scheme: "redis"
    host: "redis-master"
    port: "6379"
    db: 0

  # Deploy Temporal as a subchart (host/port auto-configured)
  temporal:
    deploy: true

  # Temporal subchart configuration (uses same PostgreSQL instance).
  # The Temporal store password is taken from your postgresql.password above —
  # the chart generates the temporal-postgresql-secret automatically.
  temporal-subchart:
    server:
      config:
        persistence:
          defaultStore: default
          visibilityStore: visibility
          datastores:
            default:
              sql:
                connectAddr: "postgresql:5432"
                databaseName: temporal
                user: <POSTGRES-USERNAME>
            visibility:
              sql:
                connectAddr: "postgresql:5432"
                databaseName: temporal_visibility
                user: <POSTGRES-USERNAME>

  config:
    llms:
      azureOpenAi:
        # One entry per Azure deployment you want LlamaCloud to use.
        # baseUrl is your Azure resource endpoint, NOT the OpenAI API.
        deployments: []
        # - model: "gpt-5.4"
        #   deploymentName: "gpt-5.4"
        #   apiKey: <AZURE-OPENAI-KEY>
        #   baseUrl: "https://<your-resource>.openai.azure.com"
        #   apiVersion: "2025-04-01-preview"
        # - model: "gpt-5.4-mini"
        #   deploymentName: "gpt-5.4-mini"
        #   apiKey: <AZURE-OPENAI-KEY>
        #   baseUrl: "https://<your-resource>.openai.azure.com"
        #   apiVersion: "2025-04-01-preview"

    frontend:
      enabled: true
    parseOcr:
      # Each OCR pod requests `nvidia.com/gpu: 1` and schedules only on GPU
      # nodes. Running OCR on CPU (gpu: false) is deprecated and will be
      # removed in a future release.
      gpu: true

    # Required. Without this the chart falls back to provider "aws" and the
    # literal default bucket names, which are almost certainly not yours.
    # See the File Storage guide for S3-compatible and non-AWS backends.
    storageBuckets:
      provider: "aws"
      parsedDocuments: <YOUR-BUCKET>
      parsedEtl: <YOUR-BUCKET>
      parsedExternalComponents: <YOUR-BUCKET>
      parsedFileParsing: <YOUR-BUCKET>
      parsedRawFile: <YOUR-BUCKET>
      parseOutput: <YOUR-BUCKET>
      parsedFileScreenshot: <YOUR-BUCKET>
      extractOutput: <YOUR-BUCKET>
      parseFileUpload: <YOUR-BUCKET>
      parseFileOutput: <YOUR-BUCKET>

    authentication:
      oidc:
        enabled: true
        discoveryUrl: "https://login.microsoftonline.com/<TENANT-ID>/v2.0/.well-known/openid-configuration"
        clientId: <CLIENT-ID>
        clientSecret: <CLIENT-SECRET>

OpenAI w/ Basic Auth

  license:
    key: <LLAMACLOUD-LICENSE-KEY>

  postgresql:
    host: "postgresql"
    port: "5432"
    database: "llamacloud"
    username: <POSTGRES-USERNAME>
    password: <POSTGRES-PASSWORD>

  mongodb:
    host: "mongodb"
    port: "27017"
    username: <MONGODB-USERNAME>
    password: <MONGODB-PASSWORD>

  rabbitmq:
    scheme: "amqp"
    host: "rabbitmq"
    port: "5672"
    username: <RABBITMQ-USERNAME>
    password: <RABBITMQ-PASSWORD>

  redis:
    scheme: "redis"
    host: "redis-master"
    port: "6379"
    db: 0

  # Deploy Temporal as a subchart (host/port auto-configured)
  temporal:
    deploy: true

  # Temporal subchart configuration (uses same PostgreSQL instance).
  # The Temporal store password is taken from your postgresql.password above —
  # the chart generates the temporal-postgresql-secret automatically.
  temporal-subchart:
    server:
      config:
        persistence:
          defaultStore: default
          visibilityStore: visibility
          datastores:
            default:
              sql:
                connectAddr: "postgresql:5432"
                databaseName: temporal
                user: <POSTGRES-USERNAME>
            visibility:
              sql:
                connectAddr: "postgresql:5432"
                databaseName: temporal_visibility
                user: <POSTGRES-USERNAME>

  config:
    llms:
      openAi:
        apiKey: <OPENAI-APIKEY>

    frontend:
      enabled: true
    parseOcr:
      # Each OCR pod requests `nvidia.com/gpu: 1` and schedules only on GPU
      # nodes. Running OCR on CPU (gpu: false) is deprecated and will be
      # removed in a future release.
      gpu: true

    # Required. Without this the chart falls back to provider "aws" and the
    # literal default bucket names, which are almost certainly not yours.
    # See the File Storage guide for S3-compatible and non-AWS backends.
    storageBuckets:
      provider: "aws"
      parsedDocuments: <YOUR-BUCKET>
      parsedEtl: <YOUR-BUCKET>
      parsedExternalComponents: <YOUR-BUCKET>
      parsedFileParsing: <YOUR-BUCKET>
      parsedRawFile: <YOUR-BUCKET>
      parseOutput: <YOUR-BUCKET>
      parsedFileScreenshot: <YOUR-BUCKET>
      extractOutput: <YOUR-BUCKET>
      parseFileUpload: <YOUR-BUCKET>
      parseFileOutput: <YOUR-BUCKET>

    authentication:
      basicAuth:
        enabled: true
        validEmailDomain: "llamaindex.ai" # required: only addresses at this domain may register
        # jwtSecret is optional. The chart generates one on first install
        # and preserves it across upgrades. Set it only to control the value.

Azure OpenAI w/ Basic Auth

  license:
    key: <LLAMACLOUD-LICENSE-KEY>

  postgresql:
    host: "postgresql"
    port: "5432"
    database: "llamacloud"
    username: <POSTGRES-USERNAME>
    password: <POSTGRES-PASSWORD>

  mongodb:
    host: "mongodb"
    port: "27017"
    username: <MONGODB-USERNAME>
    password: <MONGODB-PASSWORD>

  rabbitmq:
    scheme: "amqp"
    host: "rabbitmq"
    port: "5672"
    username: <RABBITMQ-USERNAME>
    password: <RABBITMQ-PASSWORD>

  redis:
    scheme: "redis"
    host: "redis-master"
    port: "6379"
    db: 0

  # Deploy Temporal as a subchart (host/port auto-configured)
  temporal:
    deploy: true

  # Temporal subchart configuration (uses same PostgreSQL instance).
  # The Temporal store password is taken from your postgresql.password above —
  # the chart generates the temporal-postgresql-secret automatically.
  temporal-subchart:
    server:
      config:
        persistence:
          defaultStore: default
          visibilityStore: visibility
          datastores:
            default:
              sql:
                connectAddr: "postgresql:5432"
                databaseName: temporal
                user: <POSTGRES-USERNAME>
            visibility:
              sql:
                connectAddr: "postgresql:5432"
                databaseName: temporal_visibility
                user: <POSTGRES-USERNAME>

  config:
    llms:
      azureOpenAi:
        # One entry per Azure deployment you want LlamaCloud to use.
        # baseUrl is your Azure resource endpoint, NOT the OpenAI API.
        deployments: []
        # - model: "gpt-5.4"
        #   deploymentName: "gpt-5.4"
        #   apiKey: <AZURE-OPENAI-KEY>
        #   baseUrl: "https://<your-resource>.openai.azure.com"
        #   apiVersion: "2025-04-01-preview"
        # - model: "gpt-5.4-mini"
        #   deploymentName: "gpt-5.4-mini"
        #   apiKey: <AZURE-OPENAI-KEY>
        #   baseUrl: "https://<your-resource>.openai.azure.com"
        #   apiVersion: "2025-04-01-preview"

    frontend:
      enabled: true
    parseOcr:
      # Each OCR pod requests `nvidia.com/gpu: 1` and schedules only on GPU
      # nodes. Running OCR on CPU (gpu: false) is deprecated and will be
      # removed in a future release.
      gpu: true

    # Required. Without this the chart falls back to provider "aws" and the
    # literal default bucket names, which are almost certainly not yours.
    # See the File Storage guide for S3-compatible and non-AWS backends.
    storageBuckets:
      provider: "aws"
      parsedDocuments: <YOUR-BUCKET>
      parsedEtl: <YOUR-BUCKET>
      parsedExternalComponents: <YOUR-BUCKET>
      parsedFileParsing: <YOUR-BUCKET>
      parsedRawFile: <YOUR-BUCKET>
      parseOutput: <YOUR-BUCKET>
      parsedFileScreenshot: <YOUR-BUCKET>
      extractOutput: <YOUR-BUCKET>
      parseFileUpload: <YOUR-BUCKET>
      parseFileOutput: <YOUR-BUCKET>

    authentication:
      basicAuth:
        enabled: true
        validEmailDomain: "llamaindex.ai" # required: only addresses at this domain may register
        # jwtSecret is optional. The chart generates one on first install
        # and preserves it across upgrades. Set it only to control the value.

Install the Helm chart

# Add the Helm repository
helm repo add llamaindex https://run-llama.github.io/helm-charts

# Update your local Helm chart cache
helm repo update

# Create the llamacloud namespace
kubectl create ns llamacloud

# Install the Helm chart
helm install llamacloud llamaindex/llamacloud -f values.yaml --namespace llamacloud

If you want to install a specific version of the Helm chart, you can specify the version:

helm install llamacloud llamaindex/llamacloud --version x.y.z -f values.yaml --namespace llamacloud

Validate the installation

After installation, you will see the following output:

NAME: llamacloud
LAST DEPLOYED: Tue Nov 18 10:12:03 2025
NAMESPACE: llamacloud
STATUS: deployed
REVISION: 1
TEST SUITE: None
NOTES:
Welcome to LlamaCloud!

View your deployment with the following:

  kubectl get pods -n llamacloud

To view LlamaCloud UI in your browser:
  Run the following command:

  kubectl --namespace llamacloud port-forward svc/llamacloud-web 3000:80

If you list the pods with kubectl get pods -n llamacloud, you should see the following. This is a temporal.deploy: true install, so the Temporal subchart pods appear alongside the LlamaCloud services; your backing stores show up here too when they share the namespace.

NAME                                                      READY   STATUS      RESTARTS   AGE
llamacloud-76f6dd67b6-vqbzf                               1/1     Running     0          10m
llamacloud-layout-b4887c5d9-tj9xw                         1/1     Running     0          10m
llamacloud-ocr-8ffc8bf9b-prvth                            1/1     Running     0          10m
llamacloud-operator-7b67757d88-nxlkj                      1/1     Running     0          10m
llamacloud-parse-767846797-tdkqx                          1/1     Running     0          10m
llamacloud-telemetry-9d66956fd-5d95d                      1/1     Running     0          10m
llamacloud-web-6b98cd88b-6887v                            1/1     Running     0          10m
llamacloud-worker-66cb4867fc-mjhsn                        1/1     Running     0          10m
llamacloud-temporal-parse-c89c8d9c9-zvjvv                 1/1     Running     0          10m
temporal-jobs-worker-76d6d58c77-2wlg2                     1/1     Running     0          10m
llamacloud-temporal-subchart-admintools-67cf5d57b-9cn5x   1/1     Running     0          10m
llamacloud-temporal-subchart-frontend-78cbdc8ff7-pkxgl    1/1     Running     0          10m
llamacloud-temporal-subchart-history-54cb58579f-ph42c     1/1     Running     0          10m
llamacloud-temporal-subchart-matching-858d4dcd5b-brlp2    1/1     Running     0          10m
llamacloud-temporal-subchart-web-7f6b8c9d84-5nv4k         1/1     Running     0          10m
llamacloud-temporal-subchart-worker-59b794d59-54kvd       1/1     Running     0          10m
llamacloud-temporal-subchart-schema-1-2-0-1-vrzp4         0/1     Completed   0          10m
llamacloud-temporal-search-attributes-k2czv               0/1     Completed   0          10m

The two Completed pods are one-shot jobs. The Temporal schema job provisions the temporal and temporal_visibility databases, and the search-attributes job registers the custom search attributes. They are expected to stay Completed.

Note

With config.storageBuckets.s3proxy.enabled: true, the S3-consuming pods (llamacloud, llamacloud-parse, llamacloud-worker, llamacloud-operator, llamacloud-temporal-parse, temporal-jobs-worker) run an extra S3Proxy sidecar and report 2/2 instead of 1/1.

Port forward the frontend service to access the LlamaCloud UI:

kubectl --namespace llamacloud port-forward svc/llamacloud-web 3000:80

Open your web browser and navigate to http://localhost:3000. You should see the LlamaCloud UI.

Caution

Port-forwarding is for a first smoke test only. Set ingress.host before you let anyone else use the deployment. The chart derives the deployment’s own public URL from it, and with it unset the backend falls back to the hosted product’s URL, so redirects after login and registration send users to cloud.llamaindex.ai instead of your deployment. See Ingress Configuration.

Next Steps

Choose your deployment approach based on your needs:

🌩️ Cloud-Specific Deployment Guides

Recommended for most users - Complete, opinionated guides for major cloud providers:

📋 Choose Your Cloud Provider →

• Azure: AKS + Azure-native services with Microsoft Entra ID
• AWS: EKS + AWS-native services (coming soon)
• GCP: GKE + GCP-native services (coming soon)

These guides provide end-to-end instructions using cloud-native services and enterprise authentication.

⚙️ Custom Configuration Guides

For advanced users with specific requirements or non-standard setups:

• Authentication Modes - Configure OIDC, basic auth, or custom authentication
• File Storage - Set up S3, Azure Blob, GCS, or other storage
• Database and Queues - Configure external databases and message queues
• LLM Integrations - Set up OpenAI, Azure OpenAI, Bedrock, or other LLMs
• Ingress Configuration - Load balancers, SSL, and networking
• Autoscaling Configuration - HPA and KEDA-based scaling for services
• Service Tuning - Performance and scaling configurations

Use these guides if you need custom integrations, have specific compliance requirements, or want to mix and match different services.

🚰 Index Storage Configuration

Configure where LlamaCloud Index writes the vector embeddings of your documents. The destination is set deployment-wide via config.defaultIndex.destination. See Index Configuration for the supported targets (MongoDB, Turbopuffer, PostgreSQL/pgvector, Azure AI Search) and their requirements.

More Examples and Guides

• there are many more configuration options available for each component. to see the full values.yaml specification, please refer to the values.yaml file in the helm chart repository.
• To see how common scenarios are configured, please refer to the values.yaml examples directory in the Helm chart repository.
• Similarly, we have other configuration docs available there too for more advanced configurations.