There are three ways to use Hub Executors in your project. Each has its own use case and benefits.

Use as-is#

You can use a Hub Executor as-is via Executor.from_hub():

from jina import Executor
from docarray import DocList
from docarray.documents.legacy import LegacyDocument

exec = Executor.from_hub('jinaai://jina-ai/DummyHubExecutor')
da = DocList[LegacyDocument]([LegacyDocument()])
assert da.texts == ['hello']

The Hub Executor will be pulled to your local machine and run as a native Python object. You can use a line-debugger to step in/out exec object, set breakpoints, and observe how it behaves. You can directly feed in Documents. After you build some confidence in that Executor, you can move to the next step: Using it as a part of your Flow.


Not all Executors on the Hub can be directly run in this way - some require extra dependencies. In that case, you can add .from_hub(..., install_requirements=True) to install the requirements automatically. Be careful - these dependencies may not be compatible with your local packages and may override your local development environment.


Hub Executors are cached locally on the first pull. Afterwards, they will not be updated.

To keep up-to-date with upstream, use .from_hub(..., force_update=True).

Pull only#

You can also use jina hub CLI to pull an Executor without actually using it in the Flow.

Jina and DocArray version

Independently of the Jina and DocArray version existing when the Executor was pushed to the Hub. When pulling, the Hub will try to install the Jina and DocArray version that you have installed locally in the pulled docker images.

Pull the Docker image#

jina hub pull jinaai+docker://<USERNAME>/<NAME>[:<TAG>]

You can find the Executor by running docker images. You can also indicate which version of the Executor you want to use by specifying the :<TAG>.

jina hub pull jinaai+docker://jina-ai/DummyExecutor:v1.0.0

Use in Flow as container#

Use prebuilt images from Hub in your Python code:

from jina import Flow

# You have to login for private Executor
# import hubble
# hubble.login()

f = Flow().add(uses='jinaai+docker://<USERNAME>/<NAME>[:<TAG>]')

If you do not provide a :<TAG>, it defaults to /latest.


To use a private Executor, you have to login.

import hubble



If you are a Mac user, please use host.docker.internal as your URL when you want to connect a local port from an Executor Docker container.

For example: PostgreSQLStorage will connect PostgreSQL server which was started locally. Then you must use it with:

from jina import Flow, Document

f = Flow().add(
    uses_with={'hostname': 'host.docker.internal'},
with f:
    resp ='/index', inputs=Document())

If jinaai+docker:// Executors don’t load properly or have issues during initialization, ensure you have sufficient Docker resources allocated.

Mount local volumes#

You can mount volumes into your dockerized Executor by passing a list of volumes with the volumes argument:

f = Flow().add(
    volumes=['host/path:/path/in/container', 'other/volume:/app'],


If you want your containerized Executor to operate inside one of these volumes, remember to set its workspace accordingly!

If you do not specify volumes, Jina automatically mounts a volume into the container. In this case, the volume source is your default Executor workspace, and the volume destination is /app. Additionally, automatic volume setting tries to move the Executor’s workspace into the volume destination. Depending on the default Executor workspace on your system this may not always succeed, so explicitly mounting a volume and setting a workspace is recommended.

You can disable automatic volume setting by passing f.add(..., disable_auto_volume=True).

Use in Flow via source code#

Use the source code from Executor Hub in your Python code:

from jina import Flow

f = Flow().add(uses='jinaai://<USERNAME>/<NAME>[:<TAG>]')

Set/override default parameters#

The default parameters of the published Executor may not be ideal for your use case. You can pass uses_with and uses_metas as parameters to override this:

from jina import Flow

f = Flow().add(
    uses_with={'param1': 'new_value'},
    uses_metas={'name': 'new_name'},

Platform awareness of Hub images#


As of January 10, 2023 jina hub pull is platform aware. It will automatically select Docker images based on your native CPU architecture (if available).

If you prefer a specific platform, for example, preferring AMD64 on an ARM64 machine, you can explicitly pull with --prefer-platform:


When you specify --prefer-platform you probably want to also specify --force to overwrite the existing image in local cache.


If the image you specify doesn’t support your preferred platform, it will not respect your platform preference.

jina hub pull --force --prefer-platform linux/amd64 jinaai+docker://jina-ai/DummyExecutor:v1.0.0

Pull the source code#

jina hub pull jinaai://<USERNAME>/<NAME>[:<TAG>]

List locations of local Executors#

jina hub list


To list all the Executors that are in source-code format (i.e. pulled via jinaai://), use the command jina hub list.

To list all the Executors that are in Docker format, use the command docker images.