Meta Generative AI with Weaviate
v1.39.3Weaviate's integration with Meta's API allows you to access their generative models' capabilities directly from Weaviate.
Configure a Weaviate collection to use a generative AI model with Meta. Weaviate will perform retrieval augmented generation (RAG) using the specified model and your Meta API key.
More specifically, Weaviate will perform a search, retrieve the most relevant objects, and then pass them to the Meta generative model to generate outputs.
Python client support for Meta is coming in an upcoming weaviate-client release. The snippets below do not run with the current client release.
Requirements
Weaviate configuration
Your Weaviate instance must be configured with the Meta generative AI integration (generative-meta) module.
generative-meta is an API-based module. Weaviate enables API-based modules by default, so the module is present on any v1.39.3 or later instance that has not disabled them.
Check the cluster metadata to confirm that the module is enabled on your instance.
For self-hosted users
- Follow the how-to configure modules guide to enable the module in Weaviate.
- If you set
API_BASED_MODULES_DISABLEDtotrue, or you maintain your ownENABLE_MODULESlist, addgenerative-metatoENABLE_MODULES.
API credentials
You must provide a valid Meta API key to Weaviate for this integration. See the Meta API documentation to obtain an API key.
Provide the API key to Weaviate using one of the following methods:
- Set the
META_APIKEYenvironment variable that is available to Weaviate. - Provide the API key at runtime, as shown in the examples below.
API docsIf a snippet doesn't work or you have feedback, please open a GitHub issue.
import weaviate
from weaviate.classes.init import Auth
import os
# Recommended: save sensitive data as environment variables
meta_key = os.getenv("META_APIKEY")
headers = {
"X-Meta-Api-Key": meta_key,
# "X-Meta-Baseurl": "https://api.meta.ai", # Optional; for providing a custom base URL
}
client = weaviate.connect_to_weaviate_cloud(
cluster_url=weaviate_url, # `weaviate_url`: your Weaviate URL
auth_credentials=Auth.api_key(weaviate_key), # `weaviate_key`: your Weaviate API key
headers=headers
)
# Work with Weaviate
client.close()
Configure collection
A collection's generative model integration configuration is mutable from v1.25.23, v1.26.8 and v1.27.1. See this section for details on how to update the collection configuration.
Configure a Weaviate index as follows to use a Meta generative model:
API docsIf a snippet doesn't work or you have feedback, please open a GitHub issue.
from weaviate.classes.config import Configure
client.collections.create(
"DemoCollection",
generative_config=Configure.Generative.meta()
# Additional parameters not shown
)
Select a model
Name the model in the collection configuration:
API docsIf a snippet doesn't work or you have feedback, please open a GitHub issue.
from weaviate.classes.config import Configure
client.collections.create(
"DemoCollection",
generative_config=Configure.Generative.meta(
model="muse-spark-1.2"
)
# Additional parameters not shown
)
See Available models for the names you can use and for the default. You can also override the model at query time.
Generative parameters
Configure the following generative parameters to customize the model behavior.
API docsIf a snippet doesn't work or you have feedback, please open a GitHub issue.
from weaviate.classes.config import Configure
client.collections.create(
"DemoCollection",
generative_config=Configure.Generative.meta(
# # These parameters are optional
# model="muse-spark-1.2",
# temperature=0.7, # 0.0 to 2.0
# top_p=0.9, # 0.0 to 1.0
# max_tokens=500, # 1 or higher; no upper limit
# frequency_penalty=0.0, # -2.0 to 2.0
# presence_penalty=0.0, # -2.0 to 2.0
# reasoning_effort="low", # One of "none", "minimal", "low", "medium", "high", "xhigh"
# base_url="https://api.meta.ai",
)
)
Weaviate has no default for temperature, topP, maxTokens, frequencyPenalty, presencePenalty, or reasoningEffort: any one you leave unset is omitted from the request, so Meta's own default applies.
For further details on model parameters, see the Meta API documentation.
Select a model at runtime
Aside from setting the default model provider when creating the collection, you can also override it at query time.
API docsIf a snippet doesn't work or you have feedback, please open a GitHub issue.
from weaviate.classes.generate import GenerativeConfig
collection = client.collections.use("DemoCollection")
response = collection.generate.near_text(
query="A holiday film",
limit=2,
grouped_task="Write a tweet promoting these two movies",
generative_provider=GenerativeConfig.meta(
model="muse-spark-1.2",
# # These parameters are optional
# temperature=0.7,
# top_p=0.9,
# max_tokens=500,
# frequency_penalty=0.0,
# presence_penalty=0.0,
# reasoning_effort="low",
# base_url="https://api.meta.ai",
),
# Additional parameters not shown
)
Header parameters
You can provide the API key as well as some optional parameters at runtime through additional headers in the request. The following headers are available:
X-Meta-Api-Key: The Meta API key.X-Meta-Baseurl: The base URL to use (e.g. a proxy) instead of the default Meta URL.
X-Meta-Api-Key overrides META_APIKEY, and a request with neither fails with an api key error.
X-Meta-Baseurl overrides any baseURL setting (default https://api.meta.ai) and must be an API root, because Weaviate appends /v1/chat/completions.
Provide the headers as shown in the API credentials examples above.
Retrieval augmented generation
After configuring the generative AI integration, perform RAG operations, either with the single prompt or grouped task method.
Single prompt
To generate text for each object in the search results, use the single prompt method.
The example below generates outputs for each of the n search results, where n is specified by the limit parameter.
When creating a single prompt query, use braces {} to interpolate the object properties you want Weaviate to pass on to the language model. For example, to pass on the object's title property, include {title} in the query.
API docsIf a snippet doesn't work or you have feedback, please open a GitHub issue.
collection = client.collections.use("DemoCollection")
response = collection.generate.near_text(
query="A holiday film", # The model provider integration will automatically vectorize the query
single_prompt="Translate this into French: {title}",
limit=2
)
for obj in response.objects:
print(obj.properties["title"])
print(f"Generated output: {obj.generated}") # Note that the generated output is per object
Grouped task
To generate one text for the entire set of search results, use the grouped task method.
In other words, when you have n search results, the generative model generates one output for the entire group.
API docsIf a snippet doesn't work or you have feedback, please open a GitHub issue.
collection = client.collections.use("DemoCollection")
response = collection.generate.near_text(
query="A holiday film", # The model provider integration will automatically vectorize the query
grouped_task="Write a fun tweet to promote readers to check out these films.",
limit=2
)
print(f"Generated output: {response.generative.text}") # Note that the generated output is per query
for obj in response.objects:
print(obj.properties["title"])
RAG with images
You can also supply images as a part of the input when performing retrieval augmented generation in both single prompts and grouped tasks.
API docsIf a snippet doesn't work or you have feedback, please open a GitHub issue.
import base64
import requests
from weaviate.classes.generate import GenerativeConfig, GenerativeParameters
src_img_path = "https://upload.wikimedia.org/wikipedia/commons/thumb/b/b0/Winter_forest_silver.jpg/960px-Winter_forest_silver.jpg"
base64_image = base64.b64encode(requests.get(src_img_path).content).decode('utf-8')
prompt = GenerativeParameters.grouped_task(
prompt="Which movie is closest to the image in terms of atmosphere",
images=[base64_image], # A list of base64 encoded strings of the image bytes
# image_properties=["img"], # Properties containing images in Weaviate
)
collection = client.collections.use("DemoCollection")
response = collection.generate.near_text(
query="Movies",
limit=5,
grouped_task=prompt,
generative_provider=GenerativeConfig.meta(
max_tokens=1000
),
)
# Print the source property and the generated response
for o in response.objects:
print(f"Title property: {o.properties['title']}")
print(f"Grouped task result: {response.generative.text}")
References
Available models
Weaviate passes the model name through unchecked. See Meta's documentation for the current models.
If you do not set a model, Weaviate uses muse-spark-1.2.
If long generations time out, raise MODULES_CLIENT_TIMEOUT.
Further resources
Code examples
Once the integration is configured at the collection, the data management and search operations in Weaviate work identically to any other collection. See the following model-agnostic examples:
- The How-to: Manage collections and How-to: Manage objects guides show how to perform data operations (i.e. create, read, update, delete collections and objects within them).
- The How-to: Query & Search guides show how to perform search operations (i.e. vector, keyword, hybrid) as well as retrieval augmented generation.
References
Questions and feedback
Have a question or feedback? Here's how to reach us.