Comparison
datastore.sh vs Goldsky and hosted indexers
Hosted indexing · reviewed September 2026
Short answer
Hosted indexers such as Goldsky, subgraphs, and Substreams run indexing logic that you write, sinking the output where you point it. datastore.sh sells the finished historical output as Parquet. Hosted indexing removes the servers, not the engineering: decoders, backfills, and migrations remain yours to own.
What Goldsky and hosted indexers is
With a hosted indexer you define the transformation, whether as a subgraph, a Substreams module, or a pipeline, and the platform runs it, scales it, and delivers results into a sink. It is the right shape when the output has to match an application's data model exactly and stay current.
The engineering does not disappear. Decoders track protocol upgrades, backfills re-run whenever logic changes, and every migration is yours to plan. datastore.sh sells the product of that work for the programs it covers, so the recurring obligation sits with us rather than with your team.
How the two products differ
The comparison is between product models rather than feature lists. Prices, rate limits, and chain counts change without notice, so none are quoted here.
| Dimension | datastore.sh | Goldsky and hosted indexers |
|---|---|---|
| What you receive | Partitioned Parquet files with typed schemas, manifests, and checksums | Records produced by pipeline logic you wrote |
| Who writes the decoders | We do, pinned to program versions | You do, per protocol indexed |
| Backfills | Included in the coverage window you buy | A job you run, size, and pay for on each change |
| Protocol upgrades | Handled upstream and published as new versions | A migration on your roadmap |
| Time to first data | Delivery time, with no pipeline to write | Development plus initial backfill |
| Freshness | Batch cadence, not real time | Real time, which is the core strength |
| Output shape | Standard decoded schemas, documented per table | Exactly the shape your application requires |
| Best fit | History you want to own without building for it | Live, app-specific views of protocols you choose |
When Goldsky and hosted indexers is the better choice
- You need a protocol-specific view shaped precisely to your application.
- Freshness is a product requirement, measured in seconds.
- The indexing logic is itself a differentiator worth maintaining.
When datastore.sh is the better choice
- You need history for analysis, not a live view for an application.
- Nobody on the team wants to own decoders through the next program upgrade.
- A backfill would take weeks of compute and engineering before analysis could start.
- Standard decoded schemas are sufficient, so custom pipeline logic adds no value.
Using both together
A common arrangement is an indexer for live application state and purchased archives for the historical range behind it. Buying the backfill avoids the slowest and least differentiated part of an indexing project.
Frequently asked questions
What is the difference between datastore.sh and Goldsky?
Hosted indexers such as Goldsky, subgraphs, and Substreams run indexing logic that you write, sinking the output where you point it. datastore.sh sells the finished historical output as Parquet. Hosted indexing removes the servers, not the engineering: decoders, backfills, and migrations remain yours to own.
Is buying history faster than running a backfill?
Usually, yes. A backfill over deep history costs development time, compute, and calendar time before any analysis can begin, and it repeats whenever decoding logic changes. A purchased archive arrives as files with documented schemas and checksums, so the first query runs on delivery day.
Can I keep using my indexer and still buy files?
Yes, and that combination is common. The indexer serves live application state while purchased files cover the historical range behind it. Because delivery is standard Parquet, the archive loads into the same warehouse or lake that already receives your pipeline output.
Start with the data
Skip the indexing project. Run the query.
Browse documented datasets or send the exact protocol, tables, and historical coverage your team needs.