# Openai API key fails when i connect with weaviate but works with normal chat completion api

**URL:** <https://forum.weaviate.io/t/openai-api-key-fails-when-i-connect-with-weaviate-but-works-with-normal-chat-completion-api/9964>\
**Category:** Support\
**Created:** [January 30, 2025, 9:43am UTC](https://forum.weaviate.io/t/openai-api-key-fails-when-i-connect-with-weaviate-but-works-with-normal-chat-completion-api/9964 "2025-01-30T09:43:33Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![DhanushKumar\_R](https://yyz1.discourse-cdn.com/flex027/user_avatar/forum.weaviate.io/dhanushkumar_r/32/3310_2.png) [@DhanushKumar\_R](https://forum.weaviate.io/u/DhanushKumar_R)\
**Post date:** [January 30, 2025, 9:43am UTC](https://forum.weaviate.io/t/openai-api-key-fails-when-i-connect-with-weaviate-but-works-with-normal-chat-completion-api/9964/1 "2025-01-30T09:43:33Z")

</div>

```python
# Connect to Weaviate
client = weaviate.connect_to_wcs(
    cluster_url=weaviate_url,
    auth_credentials=weaviate.auth.AuthApiKey(weaviate_key),
    headers={
        "X-OpenAI-Api-Key": openai.api_key # Replace with your OpenAI key
    }
)

# Check if Weaviate is ready
client.is_ready()

# Configure Weaviate Schema for Two Collections
import weaviate.classes.config as wc

# Schema for Credit Card Documents
client.collections.create(
    name="CreditCardDocuments",
    vectorizer_config=wc.Configure.Vectorizer.text2vec_openai(
        model="ada",
        model_version="002",
        type_="text"
    ),
    generative_config=wc.Configure.Generative.openai(
        model="gpt-4"
    ),
    properties=[
        wc.Property(name="type", data_type=wc.DataType.TEXT),
        wc.Property(name="element_id", data_type=wc.DataType.TEXT, skip_vectorization=True),
        wc.Property(name="text", data_type=wc.DataType.TEXT),
        wc.Property(name="embeddings", data_type=wc.DataType.NUMBER_ARRAY, skip_vectorization=True),
    ],
)

# Schema for HR Documents
client.collections.create(
    name="HRDocuments",
    vectorizer_config=wc.Configure.Vectorizer.text2vec_openai(
        model="ada",
        model_version="002",
        type_="text"
    ),
    generative_config=wc.Configure.Generative.openai(
        model="gpt-4"
    ),
    properties=[
        wc.Property(name="type", data_type=wc.DataType.TEXT),
        wc.Property(name="element_id", data_type=wc.DataType.TEXT, skip_vectorization=True),
        wc.Property(name="text", data_type=wc.DataType.TEXT),
        wc.Property(name="embeddings", data_type=wc.DataType.NUMBER_ARRAY, skip_vectorization=True),
    ],
)

# Define Writer for Weaviate
def get_writer(collection_name: str) -> Writer:
    return WeaviateWriter(
        connector_config=SimpleWeaviateConfig(
            access_config=WeaviateAccessConfig(api_key=weaviate_key),
            host_url=weaviate_url,
            class_name=collection_name,
        ),
        write_config=WeaviateWriteConfig(),
    )

# Ingest Data into CreditCardDocuments
credit_card_writer = get_writer("CreditCardDocuments")
credit_card_runner = S3Runner(
    processor_config=ProcessorConfig(
        verbose=True,
        output_dir="s3-output-credit-card",
        num_processes=40,
    ),
    read_config=ReadConfig(),
    partition_config=PartitionConfig(
partition_by_api=True,
        api_key="marmfWUWJpM8ncY6GQRrftfjKG7LLw", # Replace with your Unstructured API key
        partition_endpoint="", # Replace with your Unstructured API URL
    ),
    connector_config=SimpleS3Config(
        access_config=S3AccessConfig(
            key="A", # Replace with your AWS key
            secret="SI", # Replace with your AWS secret
        ),
        remote_url="s3://g/credit-card-documents", # Replace with your S3 bucket path
    ),
    chunking_config=ChunkingConfig(
        chunk_elements=True,
        chunking_strategy="by_title",
        max_characters=8192,
        combine_text_under_n_chars=1000,
    ),
    embedding_config=EmbeddingConfig(
        provider="azure",
        api_key=openai.api_key,
    ),
    writer=credit_card_writer,
    writer_kwargs={},
)

credit_card_runner.run()

# Ingest Data into HRDocuments
hr_writer = get_writer("HRDocuments")
hr_runner = S3Runner(
    processor_config=ProcessorConfig(
        verbose=True,
        output_dir="s3-output-hr",
        num_processes=40,
    ),
    read_config=ReadConfig(),
    partition_config=PartitionConfig(
        partition_by_api=True,
        api_key="mw", # Replace with your Unstructured API key
        partition_endpoint="https://api.unstructuredapp.io", # Replace with your Unstructured API URL
    ),
    connector_config=SimpleS3Config(
        access_config=S3AccessConfig(
            key="AKR", # Replace with your AWS key
            secret="nI", # Replace with your AWS secret
        ),
        remote_url="s3://gg/hr-documents", # Replace with your S3 bucket path
    ),
    chunking_config=ChunkingConfig(
        chunk_elements=True,
        chunking_strategy="by_title",
        max_characters=8192,
        combine_text_under_n_chars=1000,
    ),
    embedding_config=EmbeddingConfig(
        provider="azure",
        api_key=openai.api_key,
    ),
    writer=hr_writer,
    writer_kwargs={},
)

hr_runner.run()

# Search in CreditCardDocuments
credit_card_documents = client.collections.get("CreditCardDocuments")
credit_card_response = credit_card_documents.query.hybrid(
    query="What is the annual fee for the premium credit card?",
    alpha=0.5,
    return_properties=['text'],
    auto_limit=2
)

print("Credit Card Documents Search Results:")
for obj in credit_card_response.objects:
    print(json.dumps(obj.properties, indent=2))

# Search in HRDocuments
hr_documents = client.collections.get("HRDocuments")
hr_response = hr_documents.query.hybrid(
    query="What is the company's policy on remote work?",
    alpha=0.5,
    return_properties=['text'],
    auto_limit=2
)

print("HR Documents Search Results:")
for obj in hr_response.objects:
    print(json.dumps(obj.properties, indent=2))

```

When i try the api key with normal chat completion ,it works, but it fails when it works with weaviate

---

<div class="post-metadata">

**Author:** ![DudaNogueira](https://yyz1.discourse-cdn.com/flex027/user_avatar/forum.weaviate.io/dudanogueira/32/7846_2.png) [@DudaNogueira](https://forum.weaviate.io/u/DudaNogueira)\
**Post date:** [January 30, 2025, 1:34pm UTC](https://forum.weaviate.io/t/openai-api-key-fails-when-i-connect-with-weaviate-but-works-with-normal-chat-completion-api/9964/2 "2025-01-30T13:34:49Z")

</div>

Hi!

Can you share the questions asked in a template for a new topic? Infos like version, deployment method etc help us understand it better.

Do you see any error messages? Can you share it?

Also, it helps when you share a code we can run entirely.

I see some code (WeaviateWriter) that I don’t know it’s origin, so I can reproduce it ☹

THanks!

---

<div class="post-metadata">

**Author:** ![DhanushKumar\_R](https://yyz1.discourse-cdn.com/flex027/user_avatar/forum.weaviate.io/dhanushkumar_r/32/3310_2.png) [@DhanushKumar\_R](https://forum.weaviate.io/u/DhanushKumar_R)\
**Post date:** [January 31, 2025, 5:39am UTC](https://forum.weaviate.io/t/openai-api-key-fails-when-i-connect-with-weaviate-but-works-with-normal-chat-completion-api/9964/3 "2025-01-31T05:39:17Z")

</div>

may i have your email id because my code contains credentials and as well ,i dont feel safe to share it publicly

---

<div class="post-metadata">

**Author:** ![DudaNogueira](https://yyz1.discourse-cdn.com/flex027/user_avatar/forum.weaviate.io/dudanogueira/32/7846_2.png) [@DudaNogueira](https://forum.weaviate.io/u/DudaNogueira)\
**Post date:** [January 31, 2025, 9:11pm UTC](https://forum.weaviate.io/t/openai-api-key-fails-when-i-connect-with-weaviate-but-works-with-normal-chat-completion-api/9964/4 "2025-01-31T21:11:58Z")

</div>

You can create the code and remove any sensitive information. That will help.

From what I have seen, if you are using [unstructured-ingest](https://github.com/Unstructured-IO/unstructured-ingest), note that that integration is outdated as it is using our python v3 client 😬

```auto
    embedding_config=EmbeddingConfig(
        provider="azure",
        api_key=openai.api_key,
    ),

```

shouldn’t the provider here be openai?

Ps: I never used this code from [unstructured-ingest](https://github.com/Unstructured-IO/unstructured-ingest)