A REST API for everything Weld does

Create connections, run syncs, publish models and activate data from your own code, or embed data pipelines in your product. One documented REST API, with an OpenAPI spec for every endpoint.

Start for free Read the docs

POST/elt_syncs

# Create an ELT sync
curl -X POST \
  https://connect.weld.app/elt_syncs \
  -H "x-api-key: $API_KEY" \
  -d '{
    "connection_id": "conn_abc123",
    "destination_schema_name": "hubspot",
    "sync_interval": "0 * * * *"
  }'
▌

201Created

201 Created
{
  "id": "sync_456",
  "source_integration_id": "hubspot",
  "destination_schema_name": "hubspot",
  "sync_interval": "0 * * * *",
  "enabled": false,
  "status": "NOT_STARTED"
}
▌

Trusted by data and AI teams

Use cases

Pipelines in your product. Or behind your own scripts.

Embed data integration for your customers, or automate your own stack. Weld runs the pipelines either way.

Built for embedding

Your product starts the connection flow, configures syncs and watches their status through the API, while Weld runs the pipelines.

GET/elt_syncs/{id}/status · RUNNING

Your own connection UI

Connection bridges handle OAuth and credentials on a Weld-hosted page, and each connection's settings come back as a schema you can render yourself.

authorize_urlconnect.weld.app/authorize?token=…

Full programmatic control

Create connections, configure syncs, add or remove streams, set schedules and request full refreshes, all over REST.

POST/elt_syncs/{id}/source_streams

Orchestrate from anywhere

Create orchestrations and request runs over the API, or call it from Airflow, Prefect or any scheduler you already run.

POST/orchestrations/{id}/request_run

Watch every run

Read sync status, stream runs, reverse ETL failures and notifications, and retry failed records without opening the app.

GET/reverse_etl_failed_records/count

One key per workspace

Create an API key in Settings, send it in the x-api-key header, and it works on the workspace it belongs to.

headerx-api-key: $API_KEY

How it works

From API key to live sync. In four requests.

All requests go to https://connect.weld.app.

01
Authenticate
Create an API key in Settings → API Keys and send it with every request.
x-api-key: $API_KEY
02
Create a connection bridge
Send a redirect URI, integration ID and label, then send your user to the authorize URL that comes back.
POST /connection_bridges
03
Configure the sync
Create a sync with the connection ID, a destination schema and a cron schedule.
POST /elt_syncs
04
Add streams and enable
Pick the source streams to sync, then enable it. The pipeline runs on its schedule.
POST /elt_syncs/{id}/enable

Endpoints

84 endpoints. Every part of the pipeline.

Connections, syncs, models, orchestration and reverse ETL, with consistent responses and cursor-based pagination.

Connections

POST/connection_bridges
GET/connections
GET/connections/{id}/settings

ELT syncs and streams

POST/elt_syncs
GET/elt_syncs/{id}/status
POST/elt_streams/{id}/request_full_refresh

Transforms

POST/transforms
POST/transforms/{id}/publish
GET/transforms/{id}/versions

Orchestration

POST/orchestrations
POST/orchestrations/{id}/request_run
GET/orchestration_runs/{id}

Reverse ETL

POST/reverse_etl_syncs
POST/reverse_etl_syncs/{id}/activate
POST/reverse_etl_failed_records/retry_all

Warehouse and alerts

GET/data_warehouse
PATCH/data_warehouse
GET/notifications

See every endpoint in the API reference

endpoints in the OpenAPI spec

areas, from connections to reverse ETL

connectors you can create over the API

The platform

One warehouse. Every step of the pipeline.

Data PipelinesReplicate data from 300+ sources

Data TransformationsSQL in Weld, GitHub or dbt

Reverse ETLSync data back to your tools

AI AgentsAgentic workflows on the API

Resources

Docs, spec and a demo. Pick where to start.

Your product,
with pipelines built in.

Create an API key and run your first sync from code, on any of 300+ connectors. Free for 14 days, no credit card.

Start for free Talk to us