> For clean Markdown of any page, append `.md` to the page URL.
> For a complete documentation index, see https://docs.sarvam.ai/llms.txt.
> For full documentation content in one file, see https://docs.sarvam.ai/llms-full.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.sarvam.ai/_mcp/server.
# Get your config_id
> How to create a Config from an Extract project and get its config_id for use with the Document AI Extract API.
A `config_id` identifies a saved extraction **Config** on the Sarvam platform. Pass it in place of an inline `schema` to reuse the same extraction fields across multiple API calls without re-sending the full JSON schema each time.
## Start with an Extract project
The easiest way to create a Config is to start from an Extract project in the [Doc Agents](/docai/getting-started/overview) dashboard at [indus.sarvam.ai](https://indus.sarvam.ai/akshar/dashboard). Upload a document, describe what you want to extract, and run the project.
## Refine the schema in the Config tab
Once the project finishes processing, open it and switch to the **Config** tab. Here you can play around with the schema — add or remove fields, change types, adjust descriptions — until the extracted output matches what you need.
When you're satisfied with the schema, click **Save** to save the configuration.
## Get your config\_id
Once saved, you have two ways to get your `config_id`:
#### Copy the config\_id
After saving, the `config_id` (a UUID like `019ff577-91b7-7dde-b128-c5f418d06881`) is displayed on the Config. Copy it for use in your API call.
#### Deploy as API endpoint
Alternatively, click **Deploy as API endpoint** to get a ready-to-use curl command with the `config_id` already baked in.
## Use it in an Extract call
Pass the `config_id` where you would otherwise pass an inline `schema`:
```python
import os
from sarvamai import SarvamAI
client = SarvamAI(api_subscription_key=os.environ["SARVAM_API_KEY"])
with open("./insurance-policy.pdf", "rb") as f:
job = client.doc_ai.extract(
file=[("insurance-policy.pdf", f, "application/pdf")],
config_id="019ff577-91b7-7dde-b128-c5f418d06881",
language="en-IN",
output_format="json",
)
print("job created:", job.job_id, "| status:", job.status)
```
**Provide exactly one of `schema` or `config_id`.** Sending both, or neither, returns a `400`.