> ## Documentation Index
> Fetch the complete documentation index at: https://docs.blockworks.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Overview

> Blockworks datasets delivered natively into your own data warehouse: SQL access at warehouse scale, no ETL.

A Blockworks Datashare delivers the same warehouse-modeled datasets that back the
[Data API](/getting-started/data-api) directly into your own data warehouse. You query
them as if they were your own tables: no pipeline to build, no data to copy, and no rows
to page through.

Snowflake is available today, through Snowflake Secure Data Sharing.

<CardGroup cols={1}>
  <Card title="Snowflake" icon="snowflake" href="/datashare/snowflake/getting-access">
    Mount the Blockworks share in your Snowflake account and query it with ordinary SQL.
  </Card>
</CardGroup>

## Why a datashare

<CardGroup cols={2}>
  <Card title="No ETL" icon="refresh">
    Data is shared in place. You point your warehouse at the share and query it. Nothing
    is copied into your account and nothing needs syncing.
  </Card>

  <Card title="Full SQL" icon="database">
    Joins, window functions, aggregations, and CTEs against the full dataset, on your own
    compute.
  </Card>

  <Card title="Joins with your data" icon="link">
    Combine Blockworks datasets with your own tables in a single query: positions,
    customers, internal metrics.
  </Card>

  <Card title="Your governance" icon="shield-check">
    Access is controlled through your own roles and grants, and queries run under your
    account's existing policies.
  </Card>
</CardGroup>

## Datashare or Data API?

Both surface the same underlying data. The right choice is about access pattern, not
content.

|                        | Data API                                              | Datashare                                 |
| ---------------------- | ----------------------------------------------------- | ----------------------------------------- |
| **Interface**          | REST over HTTPS                                       | SQL in your warehouse                     |
| **Best for**           | Applications, dashboards, agents, per-request lookups | Analytics, research, backtests, bulk work |
| **Shape of work**      | Filtered, paginated reads                             | Arbitrary queries over full history       |
| **Joins to your data** | In your application                                   | In the warehouse                          |
| **Compute**            | Ours                                                  | Yours                                     |
| **Latency**            | Millisecond responses, edge-cacheable                 | Warehouse query time                      |

<Note>
  These are complementary. A common pattern is research and model development against the
  datashare, then serving the resulting product from the Data API.
</Note>

Reach for a Datashare when you want to scan or aggregate more data than a REST endpoint
should reasonably return, when your analysis is fundamentally a SQL problem, or when the
answer requires joining Blockworks data to data of your own.

Reach for the [Data API](/getting-started/data-api) when you are serving a request in an
application, when you need low latency, or when a warehouse is not part of your stack.

## Where the data comes from

Blockworks datasets are modeled in a Snowflake warehouse and version-controlled with dbt
by the Blockworks data team. That warehouse is the single source
of truth: the Data API serves a fast mirror of these models, and a datashare exposes them
natively. Both reflect the same modeling and the same definitions.

## Other platforms

Snowflake is the platform we deliver today. If you work in Databricks, BigQuery, or
another warehouse, tell us. Delivery to additional platforms is arranged case by case,
and knowing what you need shapes what we build next.

<Card title="Contact Blockworks" icon="mail" href="https://blockworks.com/contact">
  Get in touch at [blockworks.com/contact](https://blockworks.com/contact) to discuss
  datashare access or a platform we do not yet list.
</Card>