> ## Documentation Index
> Fetch the complete documentation index at: https://portkey-docs-add-third-party-integration-issues-fixes.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Supabase

> Generate embeddings with Portkey and store them in Supabase pgvector

Supabase provides Postgres with pgvector for AI applications. With Portkey, generate embeddings using 1600+ models and store them in Supabase for efficient retrieval.

## Prerequisites

* Supabase project API key
* [Portkey API key](https://app.portkey.ai/api-keys)

## Setup

### 1. Install Dependencies

```sh theme={null}
pip install portkey-ai supabase
```

### 2. Prepare Database

<Steps>
  <Step title="Create Supabase Project">
    Go to [Supabase](https://supabase.com/dashboard/sign-in) and create a new project.
  </Step>

  <Step title="Enable pgvector">
    In **Database → Extensions**, enable `pgvector`. Or run:

    ```sql theme={null}
    create extension vector;
    ```
  </Step>

  <Step title="Create Documents Table">
    Create a table for documents and embeddings. OpenAI's `text-embedding-ada-002` outputs 1536 dimensions:

    ```sql theme={null}
    create table documents (
      id bigserial primary key,
      content text,
      embedding vector(1536)
    );
    ```
  </Step>
</Steps>

### 3. Configure Clients

```python theme={null}
from portkey_ai import Portkey
from supabase import create_client, Client

# Supabase
supabase: Client = create_client(
    "YOUR_SUPABASE_PROJECT_URL",
    "YOUR_SUPABASE_API_KEY"
)

# Portkey
client = Portkey(
    api_key="YOUR_PORTKEY_API_KEY",
    provider="@openai-prod"
)
```

## Generate & Store Embeddings

```python theme={null}
# Generate embedding
response = client.embeddings.create(
    model="text-embedding-ada-002",
    input="The food was delicious and the waiter..."
)

embedding = response.data[0].embedding

# Store in Supabase
result = supabase.table('documents').insert({
    "content": "The food was delicious and the waiter...",
    "embedding": embedding
}).execute()
```

## Switch Providers

Change the provider to use different embedding models:

```python theme={null}
# Cohere embeddings (1024 dimensions)
client = Portkey(
    api_key="YOUR_PORTKEY_API_KEY",
    provider="@cohere-prod"
)

response = client.embeddings.create(
    model="embed-english-v3.0",
    input_type="search_query",
    input="The food was delicious and the waiter..."
)
```

<Note>
  Cohere's `embed-english-v3.0` outputs 1024 dimensions. Create a separate table with `vector(1024)` for Cohere embeddings.
</Note>

## Next Steps

<CardGroup cols={2}>
  <Card title="Model Catalog" icon="list" href="/product/model-catalog">
    Set up embedding providers
  </Card>

  <Card title="Supported Providers" icon="plug" href="/integrations/llms">
    See all 1600+ supported models
  </Card>
</CardGroup>