Softwr

Databases · head to head

Dgraph vs turbopuffer

Dgraph logo

Dgraph

Databases

Apache 2.0 distributed graph database written in Go, maintained by Hypermode, queried through GraphQL or its own DQL language.

From
Free
Rated
-
Full reviewWebsite
turbopuffer logo

turbopuffer

Databases

Closed-source vector and full-text search service built directly on object storage, with cold queries measured in seconds rather than milliseconds.

From
$16/month
Rated
-
Full reviewWebsite

The short version

  • Only Dgraph has a free tier, so it costs nothing to try first.
  • Each has a real cost: Dgraph sharding is by predicate, so a single very hot predicate lives entirely in one Raft group and cannot be split further; adding nodes does not relieve it and the fix is a data model change.; turbopuffer a cold namespace pays object storage latency on the first query, with a documented p90 around 1,214 ms on a million documents, so any interactive search box needs the data kept warm or the user waits about a second.
  • They diverge on capability: Dgraph covers Apache 2.0 licence, turbopuffer covers Object storage architecture.
  • Prices and features above were last checked on 30 August 2026.

Where they differ

Only the attributes on which Dgraph and turbopuffer actually diverge.

Attributes where Dgraph and turbopuffer differ
AttributeDgraphturbopuffer
Starting priceFree$16/month
Pricing modelfreemiumsubscription
Free tierYesNo
PlatformsLinux, Mac, Docker, WebWeb
Founded2016Unknown

Identical on both: user rating (Not yet rated), category (Databases).

What each one covers

Drawn from each product's published feature list. An absence here means we hold no record of it - not that the product lacks it.

Only in Dgraph

  • Apache 2.0 licence
  • Generated GraphQL API
  • DQL query language
  • Predicate sharding
  • Raft replication
  • Distributed ACID transactions
  • Written in Go
  • Full-text and geo indexing

Only in turbopuffer

  • Object storage architecture
  • Namespaces
  • Vector search
  • Full-text search
  • Attribute filtering
  • Documented limits
  • Configurable consistency
  • Durable writes

What people use each for

The jobs each tool is most often brought in to do.

Dgraph

  • An application whose core data is a graph, such as permissions, social connections or product relationships, where the frontend already consumes GraphQLnot turbopuffer
  • A knowledge graph that outgrew a single machine and needs storage distributed across nodes without a per-core licence negotiationnot turbopuffer
  • Recommendation or fraud-detection features that traverse several hops at request time, where a relational join chain has become the bottlenecknot turbopuffer
  • Teams that want a graph database they can read, fork and self-host under a permissive licence rather than a source-available onenot turbopuffer

turbopuffer

  • A product with one search index per customer and thousands of customers, most of whose data is idle on any given daynot Dgraph
  • Very large corpora where holding every vector in memory is the dominant cost and occasional cold-query latency is acceptablenot Dgraph
  • Hybrid retrieval combining BM25 and vector search where running and synchronising two separate systems is the problem being solvednot Dgraph
  • Retrieval for agent and assistant products where indexes are created and destroyed frequently and per-index overhead must be near zeronot Dgraph

Where each one falls short

Documented limitations, not opinions. Every one is a constraint you would hit in normal use.

Dgraph

  • Sharding is by predicate, so a single very hot predicate lives entirely in one Raft group and cannot be split further; adding nodes does not relieve it and the fix is a data model change.
  • Stewardship passed from Dgraph Labs to Hypermode after the original company stopped operating independently, so anyone building on it is betting on a second custodian rather than on the original team's roadmap.
  • The GraphQL layer is generated and opinionated, so anything it does not express drops you into DQL, which is a second language your team must learn and which no other database speaks.
  • The community is a fraction of Neo4j's, so operational answers, tuning experience, hiring and third-party tooling are all thinner, and unusual failure modes in a Zero and Alpha cluster leave you reading source rather than a forum.
  • There is no portable graph standard to migrate to; Cypher, Gremlin and SPARQL are all different query models, so the schema and every query is a rewrite if you later leave, and that cost grows with the application.

turbopuffer

  • A cold namespace pays object storage latency on the first query, with a documented p90 around 1,214 ms on a million documents, so any interactive search box needs the data kept warm or the user waits about a second.
  • Queries are eventually consistent by default, and after roughly 128 MiB of outstanding writes new data is invisible until indexed, which the vendor puts at tens of seconds for small namespaces and tens of minutes for large ones, so a bulk re-index is not immediately queryable.
  • It is closed source with no community edition, so single-tenant or bring-your-own-cloud deployment is a commercial negotiation rather than a deployment choice, and there is no path to running it yourself if the relationship ends.
  • Per-namespace ceilings, roughly 10,000 writes per second, 32 MB/s and 500 million documents per shard, mean a single enormous index has to be sharded across namespaces by your application rather than by the service.
  • It is a search engine, not a database: there are no joins, no cross-document transactions and no SQL, so it sits beside a primary datastore and keeping the two in step is work that belongs to you.

Pricing, plan by plan

Dgraph

Free
  • CommunityFree
    • Native GraphQL
    • Graph queries
    • Full-text search
  • Cloud$39/month
    • Managed service
    • Auto-scaling
    • Enterprise support

turbopuffer

$16/month
  • Launch$16/month
    • All database features
    • Multi-tenancy deployment
    • SOC2 & GDPR-ready DPA
  • Scale$256/month
    • Everything in Launch
    • HIPAA-ready BAA
    • Single Sign-On (SSO)
  • Enterprise$4096/month
    • Everything in Scale
    • Single-tenancy & BYOC deployment options
    • Private networking

Which should you pick?

Choose Dgraph if

  • You need apache 2.0 licence.
  • You want to start without paying.
  • You work on Linux, Mac, Docker, Web.
  • You also want generated graphql api.
Dgraph review

Choose turbopuffer if

  • You need object storage architecture.
  • You also want namespaces.
turbopuffer review

Questions people ask

Is Dgraph or turbopuffer better?
Neither clearly leads. Dgraph starts at Free and turbopuffer at $16/month, and user ratings are close enough to be indistinguishable. Choose on capability and platform support.
Which is cheaper, Dgraph or turbopuffer?
Dgraph has a free tier; the other does not. Paid plans start at Free for Dgraph and $16/month for turbopuffer.
Does Dgraph or turbopuffer run on more platforms?
Dgraph runs on Linux, Mac, Docker, Web. turbopuffer runs on Web.
Can I use Dgraph for free?
Yes. Dgraph has a free tier, so you can try it without paying. turbopuffer starts at $16/month.
What is Dgraph best used for?
Dgraph is most often used for an application whose core data is a graph, such as permissions, social connections or product relationships, where the frontend already consumes graphql, a knowledge graph that outgrew a single machine and needs storage distributed across nodes without a per-core licence negotiation, recommendation or fraud-detection features that traverse several hops at request time, where a relational join chain has become the bottleneck, teams that want a graph database they can read, fork and self-host under a permissive licence rather than a source-available one. Of those, an application whose core data is a graph, such as permissions, social connections or product relationships, where the frontend already consumes graphql and a knowledge graph that outgrew a single machine and needs storage distributed across nodes without a per-core licence negotiation are not what turbopuffer is typically brought in for.
What can Dgraph do that turbopuffer cannot?
Dgraph covers Apache 2.0 licence, Generated GraphQL API, DQL query language, Predicate sharding. turbopuffer covers Object storage architecture, Namespaces, Vector search, Full-text search.

Answered from the vendors’ own pages

Dgraph: Is Dgraph open source?

Yes. The current repository is Apache 2.0, which is a permissive OSI licence, and the project is at v25 under Hypermode's maintenance.

turbopuffer: Can I self-host turbopuffer?

There is no open source or community edition. Single-tenant and bring-your-own-cloud deployments exist as commercial arrangements, but there is no way to run it independently of the vendor.

Dgraph: Is it really GraphQL?

It serves a generated GraphQL API, which is real GraphQL for clients. Its native language, DQL, resembles GraphQL syntactically but is Dgraph's own language and is not the GraphQL specification.

turbopuffer: How fast is it really?

Warm queries perform comparably to in-memory search engines. Cold queries, where data is not cached, have a documented p90 around 1,214 ms on a million documents. Write p90 is around 248 ms for a 512 KB upsert because writes go straight to object storage.

Dgraph: How does it compare to Neo4j?

Neo4j has the larger ecosystem, Cypher, and far more operational precedent. Dgraph distributes storage across nodes by default and gives you a GraphQL endpoint without writing resolvers. The choice usually turns on whether you need horizontal scale and a GraphQL surface more than you need ecosystem depth.

turbopuffer: Is it consistent?

Eventually consistent by default, with the vendor reporting that over 99.8% of queries return consistent data. Strong consistency can be requested per query at a latency cost. Large write bursts have a longer visibility delay while indexing catches up.

Dgraph: What does a production cluster look like?

At minimum a set of Zero nodes for coordination and a replicated set of Alpha nodes for data, typically three of each for fault tolerance, which is a meaningfully larger operational footprint than a single graph server.

turbopuffer: What is it best at?

Large numbers of namespaces where most are idle. The architecture makes cold data cheap to keep, which is exactly the shape of a multi-tenant product with a long tail of inactive customers.

Dgraph: Who maintains it now?

Hypermode. Dgraph Labs, the original company, no longer operates it, and that change of custodian is the main non-technical risk to weigh.

turbopuffer: What are the hard limits?

Up to 128 billion documents and 256 TB per namespace, 500 million documents per shard, 64 MiB per document, 10,752 dense vector dimensions, roughly 10,000 writes per second per namespace and a maximum result set of 10,000.

Share

Related pages

More on Dgraph

More on turbopuffer

Keep looking

Other head to heads