modal.Dict

class Dict(modal.object.Object)

Distributed dictionary for storage in Modal apps.

Dict contents can be essentially any object so long as they can be serialized by cloudpickle. This includes other Modal objects. If writing and reading in different environments (eg., writing locally and reading remotely), it’s necessary to have the library defining the data type installed, with compatible versions, on both sides. Additionally, cloudpickle serialization is not guaranteed to be deterministic, so it is generally recommended to use primitive types for keys.

Lifetime of a Dict and its items

An individual Dict entry will expire after 7 days of inactivity (no reads or writes). The Dict entries are written to durable storage.

Legacy Dicts (created before 2025-05-20) will still have entries expire 30 days after being last added. Additionally, contents are stored in memory on the Modal server and could be lost due to unexpected server restarts. Eventually, these Dicts will be fully sunset.

Usage

from modal import Dict

my_dict = Dict.from_name("my-persisted_dict", create_if_missing=True)

my_dict["some key"] = "some value"
my_dict[123] = 456

assert my_dict["some key"] == "some value"
assert my_dict[123] == 456

The Dict class offers a few methods for operations that are usually accomplished in Python with operators, such as Dict.put and Dict.contains. The advantage of these methods is that they can be safely called in an asynchronous context by using the .aio suffix on the method, whereas their operator-based analogues will always run synchronously and block the event loop.

For more examples, see the guide.

hydrate 

hydrate(self, client=None)

Synchronize the local object with its identity on the Modal server.

It is rarely necessary to call this method explicitly, as most operations will lazily hydrate when needed. The main use case is when you need to access object metadata, such as its ID.

Added in v0.72.39: This method replaces the deprecated .resolve() method.

objects 

objects: DictManager

Namespace with methods for managing named Dict objects.

objects.create 

create(self, name, *, allow_existing=False, environment_name=None, client=None)

Create a new named Dict in the workspace environment.

This does not return a local handle; use modal.Dict.from_name to look up the Dict after creation.

Added in v1.1.2.

Parameters

Usage

modal.Dict.objects.create("my-dict")

Dicts will be created in the active environment, or another one can be specified:

modal.Dict.objects.create("my-dict", environment_name="dev")

By default, an error is raised if the Dict already exists; allow_existing=True makes that case a no-op:

modal.Dict.objects.create("my-dict", allow_existing=True)

Note that this method does not return a local instance of the Dict. You can use modal.Dict.from_name to perform a lookup after creation.

objects.list 

list(self, *, max_objects=None, created_before=None, environment_name="",
    client=None)

List named Dicts in the workspace environment as hydrated handles.

Results are ordered newest to oldest. By default, all matching Dicts are returned.

Added in v1.1.2.

Parameters

Returns

Hydrated Dict objects for each named Dict in the listing.

Usage

dicts = modal.Dict.objects.list()
print([d.name for d in dicts])

Dicts will be retrieved from the active environment, or another one can be specified:

dev_dicts = modal.Dict.objects.list(environment_name="dev")

By default, all named Dicts are returned, newest to oldest. It’s also possible to limit the number of results and to filter by creation date:

dicts = modal.Dict.objects.list(max_objects=10, created_before="2025-01-01")

objects.delete 

delete(self, name, *, allow_missing=False, environment_name=None, client=None)

Delete a named Dict entirely (not a single key).

Deletion is irreversible and affects any Apps using this Dict.

Added in v1.1.2.

Parameters

Usage

await modal.Dict.objects.delete("my-dict")

Dicts will be deleted from the active environment, or another one can be specified:

await modal.Dict.objects.delete("my-dict", environment_name="dev")

name 

name(self)

Name of the Dict object.

Usage

d = modal.Dict.from_name("my-dict")
print(d.name)

ephemeral 

ephemeral(cls, *, client=None, environment_name=None)

Create an anonymous Dict that exists for the duration of the context manager.

Parameters

Usage

from modal import Dict

with Dict.ephemeral() as d:
    d["foo"] = "bar"

async with Dict.ephemeral() as d:
    await d.put.aio("foo", "bar")

from_name 

from_name(name, *, environment_name=None, create_if_missing=False, client=None)

Reference a named Dict, optionally creating it on the server first.

Hydration is lazy: metadata is fetched from Modal the first time the handle is used.

Parameters

Returns

A Dict handle (possibly not yet hydrated).

Usage

d = modal.Dict.from_name("my-dict", create_if_missing=True)
d[123] = 456

from_id 

from_id(dict_id, client=None)

Construct a Dict from an id and look up the Dict metadata.

This is a lazy method that defers hydrating the local object with metadata from Modal servers until the first time it is actually used.

The ID of a Dict object can be accessed using .object_id.

Parameters

Returns

A Dict handle (possibly not yet hydrated).

Usage

@app.function()
def my_worker(dict_id: str):
    d = modal.Dict.from_id(dict_id)
    d["key"] = "Hello from remote function!"

with modal.Dict.ephemeral() as d:
    my_worker.remote(d.object_id)
    print(d["key"])  # "Hello from remote function!"

info 

info(self)

Return information about the Dict object.

clear 

clear(self)

Remove all items from the Dict.

get 

get(self, key, default=None)

Get the value associated with a key.

Returns default if key does not exist.

contains 

contains(self, key)

Return if a key is present.

len 

len(self)

Return the length of the Dict.

Note: This is an expensive operation and will return at most 100,000.

update 

update(self, other=None, **kwargs)

Update the Dict with additional items.

put 

put(self, key, value, *, skip_if_exists=False)

Add a specific key-value pair to the Dict.

Returns True if the key-value pair was added and False if it wasn’t because the key already existed and skip_if_exists was set.

pop 

pop(self, key, default=_NO_DEFAULT)

Remove a key from the Dict, returning the value if it exists.

If key is not found, return default if provided, otherwise raise KeyError.

keys 

keys(self)

Return an iterator over the keys in this Dict.

Note that (unlike with Python dicts) the return value is a simple iterator, and results are unordered.

values 

values(self)

Return an iterator over the values in this Dict.

Note that (unlike with Python dicts) the return value is a simple iterator, and results are unordered.

items 

items(self)

Return an iterator over the (key, value) tuples in this Dict.

Note that (unlike with Python dicts) the return value is a simple iterator, and results are unordered.