GraphForge vs. SurrealDB
Can one database cover graph, vector, document, and embedded workloads?
SurrealDB is a Rust multimodel database: documents, record links, RELATE edge tables, vectors, full-text, and geospatial data, queried with SurrealQL or, since 3.2, ISO GQL. It can run embedded, in a browser through WebAssembly, as a single server, or as a cluster. It has no graph algorithm library beyond path modifiers. GraphForge is graph-only, but ships 94 algorithms and an investigation history inside the same embedded engine.
| GraphForge | SurrealDB | |
|---|---|---|
| Model | Embedded property graph | Multimodel document-graph |
| Version | 0.5.2 stable · 0.6.0-rc.52 | 3.3.4 (2026-10-13)18 |
| License | Apache-2.0 (since 0.5.0) | Business Source License 1.1 (converts to Apache-2.0 on 2030-01-01); commercial Enterprise licence |
| Status | Actively maintained2 | Actively maintained18 |
On this page
Should I choose SurrealDB or GraphForge?
SurrealDB for one database across documents, graph, and vectors; GraphForge for graph algorithms and analysis.
Choose GraphForge when
- You need centrality, community detection, similarity, or flow algorithms, which SurrealDB does not ship.
- You want results as Arrow tables for dataframes and notebooks.
- You want an Apache-2.0 license rather than BSL 1.1.
Choose SurrealDB when
- An application needs one database for documents, graph edges, vectors, and geospatial data.
- The same engine must run in a browser, on a device, on a server, and in a managed cluster.
- You want time-travel queries, using versioned RocksDB or SurrealKV, alongside application data.
What can each one do?
20 of 26 capabilities differ. For 7, SurrealDB's docs did not confirm support, so they are marked Not established rather than No.
Each capability has a status and a source. “Not established” means we could not confirm the feature from SurrealDB's primary documentation. It does not mean SurrealDB lacks it.
| Capability | GraphForge | SurrealDB |
|---|---|---|
| Storage | ||
| Property graph model | Shipped Labeled property graph with typed properties. 1 | Yes RELATE creates edges as standalone edge-table records with in/out fields and arbitrary properties. 28 |
| Document model | Not offered No document-store model; properties accept nested, heterogeneous values. 1 | Yes Document records are a core model alongside graph, vector, time-series and geospatial. 27 |
| Relational tables | Not offered Not a relational database; results arrive as Arrow tables. 1 | Partial Tables, schemas and record links, but no SQL; 3.3 Postgres wire protocol carries SurrealQL/GQL. 18 |
| Key-value access | Not offered No key-value API; data is a property graph. 1 | Not established Built on transactional KV storage engines, but no user-facing key-value model documented. |
| Columnar storage | Shipped Durable projects store graph data as Arrow and Parquet. 4 | Not established Not documented. |
| Vector search | Shipped Exact cosine retrieval over embeddings you supply; no ANN index. 5 | Yes HNSW (in-memory) and DISKANN (on-disk, since 3.1.0) vector indexes plus brute-force search. 23 |
| Full-text search | Shipped Tantivy-backed BM25; hybrid text and vector retrieval through find(). 5 | Yes FULLTEXT ANALYZER indexes with optional BM25 ranking and highlights. 23 |
| Geospatial | Not offered No spatial types or indexes are documented. | Yes Geospatial is listed as a supported data model alongside document, graph and vector. 27 |
| Execution | ||
| Transactional workloads (OLTP) | Partial Local analytical scope; durable atomic writes with queued and optimistic writer modes. 6 | Yes Every transaction runs under snapshot isolation with write-conflict detection on commit. 19 |
| Analytical workloads (OLAP) | Shipped Analytical verbs and Cypher over a local columnar project. 7 | Not established No dedicated analytical engine documented. |
| Graph traversal | Shipped Cypher MATCH patterns plus BFS, DFS, Dijkstra, A*, and reachability. 8 | Yes Arrow syntax ->, <-, <-> traverses edge tables from records. 28 |
| Variable-length paths | Shipped Bounded and unbounded patterns and shortestPath(). 8 | Yes Recursive paths {n}, {m..n}, {..} over links or edges, depth 1 to 256. 21 |
| Graph algorithms | Shipped 94 native algorithms behind analysis verbs, not Cypher CALL procedures. 37 | Partial Only path built-ins (+path, +collect, +shortest; GQL SHORTEST); no centrality/community library. 2126 |
| Incremental computation | Not offered Algorithms recompute on request; no incremental mode is documented. 7 | Yes DEFINE TABLE ... AS SELECT views update incrementally on writes to the source table. 24 |
| Deployment | ||
| Embedded / in-process | Shipped In-process library; no server, daemon, or connection string. 9 | Yes In-process embedding from Rust, .NET, Go, Python and JavaScript SDKs. 25 |
| Server mode | Planned Optional remote peer extension planned after 0.6.0. 11 | Yes surreal start runs a server on memory, RocksDB, SurrealKV or TiKV storage. 20 |
| WebAssembly | Not planned A browser WebAssembly engine was declined in ADR 0027. 12 | Yes @surrealdb/wasm runs the engine in browsers with mem:// or IndexedDB persistence. 25 |
| Distributed cluster | Not planned Distributed clusters and HA replication are out of scope. 9 | Paid edition Production multi-node HA on Cloud Scale or Enterprise; Community TiKV only for experimentation. 2019 |
| Managed cloud | Not planned No hosted compute. The graphforge.sh Hub distributes repositories; execution stays local. 13 | Yes SurrealDB Cloud: single-node Start instances to multi-node Scale clusters. 19 |
| Investigation | ||
| Branching | Experimental Research versions, branches, slices, proposals, and forks ship in 0.6.0 release candidates. 14 | Not established Not documented. |
| Snapshots / time travel | Shipped Named checkpoints you can diff and revert, plus transaction-time snapshots. 3 | Partial SELECT ... VERSION <datetime> needs versioned RocksDB or beta SurrealKV; not in-memory engine. 22 |
| Immutable history | Shipped Knowledge ledger records immutable assertions and evidence lineage. 16 | Not established Versioning has configurable retention; no immutable commit history documented. |
| Provenance | Shipped Provenance and knowledge ledger with evidence lineage. 16 | Not established Changefeeds show per-record diffs, but no provenance model documented. |
| Competing hypotheses | Shipped Immutable hypothesis groups with membership and selection events. 3 | Not established |
| Operations | ||
| ACID transactions | Partial Acknowledged-durable atomic publication; docs do not claim general ACID or serializable isolation. 6 | Yes Transactional storage contract with snapshot isolation; SELECT ... FOR UPDATE added in 3.3. 1918 |
| Replication | Not planned HA replication is out of scope. 9 | Paid edition Replication/fault tolerance come from distributed storage on Cloud Scale or Enterprise. 19 |
Showing 26 of 26 capabilities.
How is each one built?
SurrealDB: multimodel document-graph. GraphForge: an embedded property graph library in Rust, persisted as Parquet.
GraphForge
GraphForge is a Rust library that runs inside your process: no daemon, server, or connection string. Projects persist to a local directory as Arrow and Parquet data plus JSON metadata, and every query and analysis verb returns an Apache Arrow table. Algorithms run in the same native engine as Cypher and are called as API verbs (rank, cluster, paths, analyze, similar), not as Cypher procedures.
SurrealDB
SurrealDB separates a Rust query engine from pluggable storage: RocksDB, SurrealKV (beta), the in-memory SurrealMX, and IndexedDB in browsers. Graph edges are standalone tables with in and out fields. Recursive path syntax walks links up to 256 hops with +path, +collect, and +shortest.
SurrealDB spreads one engine across many data models and deployment targets. GraphForge goes deeper on one model: the graph, its algorithms, and its investigation record.
What the documentation says about SurrealDB
- SurrealDB 3.x is licensed under BSL 1.1 with an Additional Use Grant that forbids offering it as a Database Service, changing to Apache-2.0 on 2030-01-01.17
- SurrealDB separates the query engine (compute) from a pluggable storage layer: RocksDB, SurrealKV (beta), in-memory SurrealMX, and IndexedDB in browsers.19
- FoundationDB support was removed in SurrealDB 3.0.0; fdb:// paths are only accepted by 2.x.20
- Graph edges created with RELATE are standalone tables with required in and out fields that can hold other properties.28
- Recursive path syntax traverses record links or graph edges to a depth from 1 to 256, with +path, +collect and +shortest modifiers.21
- ISO GQL (ISO/IEC 39075) support arrived in 3.2.0 behind an experimental flag and is enabled by default from 3.3.0.26
- SurrealDB 3.3.0 (2026-09-24) added an opt-in Postgres wire protocol so Postgres clients can run SurrealQL or GQL.18
- The engine can be embedded in-process from Rust, .NET, Go, Python and JavaScript, including browsers via WebAssembly.25
- Pre-computed table views defined with DEFINE TABLE ... AS SELECT are updated incrementally when the source table is written.24
Which one is faster?
Not measured yet. No head-to-head run has been published, so this page shows no performance chart.
Charts appear only after a reproducible run exists: pinned versions, disclosed hardware, generated data, raw results, and scripts.
No comparative results yet
The suites below apply to GraphForge and SurrealDB. Each will be reported separately per benchmark class, never merged into one leaderboard.
Ingestion and writes
| Suite | Workloads | Metrics | Status |
|---|---|---|---|
| INGEST | CSV, JSONL, Parquet, graph edge lists | Records/s, MB/s, total elapsed | Not run |
| MUTATE | Insert, update, delete, upsert | Ops/s, p50/p95/p99 latency | Not run |
Query, analytics, and system
| Suite | Workloads | Metrics | Status |
|---|---|---|---|
| LOOKUP | ID lookup, indexed property filter | p50/p95/p99 latency | Not run |
| TRAVERSE | BFS, k-hop, variable-length path | Latency, visited edges/s | Not run |
| JOIN | Multi-relation pattern matching | Latency, rows/s | Not run |
| ALGO | PageRank, WCC, SCC, Louvain, SSSP | Time, memory, correctness | Not run |
| HYBRID | Vector + graph + metadata filters | Recall@k, latency, throughput | Not run |
| EMBED | Startup, load, reopen, binary size | ms, MiB, disk footprint | Not run |
Which graph algorithms ship?
GraphForge: 94 native algorithms. SurrealDB: 4 named in its docs, delivered as bundled.
Algorithms are listed by name, as documented. Counts of extension modules or third-party libraries are not presented as native coverage.
GraphForge
94 Rust algorithms run in the engine behind five analysis verbs (rank, cluster, paths, analyze, similar) and return Arrow tables.37
A selection of the catalog:
- PageRank
- ArticleRank
- Betweenness (Brandes)
- Closeness
- Harmonic closeness
- Eigenvector
- HITS
- CELF influence maximization
- Triangle count
- K-core decomposition
- Louvain
- Leiden
- Label propagation
- Infomap
- Walktrap
- Girvan-Newman
- Weakly connected components
- Strongly connected components
- Dijkstra
- A*
- Bellman-Ford
- Yen's k-shortest paths
- Max flow / min cut
- Steiner tree
- Minimum spanning tree
- Topological sort
- Node2Vec
- GraphSAGE
- Fast random projection
- Node similarity (Jaccard)
SurrealDB
No graph-algorithm library or graph function module; path-finding is offered as recursive-path modifiers in SurrealQL and SHORTEST path patterns in GQL.2126
Delivered as: bundled
- +shortest (shortest path to a record)
- +path (collect all walked paths)
- +collect (collect unique reachable nodes)
- GQL SHORTEST / ALL SHORTEST path search
Where can each one run?
GraphForge runs in-process only. SurrealDB: embedded yes, server yes, cluster paid edition.
| Mode | GraphForge | SurrealDB |
|---|---|---|
| Embedded / in-process | Shipped In-process library; no server, daemon, or connection string. 9 | Yes In-process embedding from Rust, .NET, Go, Python and JavaScript SDKs. 25 |
| Server mode | Planned Optional remote peer extension planned after 0.6.0. 11 | Yes surreal start runs a server on memory, RocksDB, SurrealKV or TiKV storage. 20 |
| WebAssembly | Not planned A browser WebAssembly engine was declined in ADR 0027. 12 | Yes @surrealdb/wasm runs the engine in browsers with mem:// or IndexedDB persistence. 25 |
| Distributed cluster | Not planned Distributed clusters and HA replication are out of scope. 9 | Paid edition Production multi-node HA on Cloud Scale or Enterprise; Community TiKV only for experimentation. 2019 |
| Managed cloud | Not planned No hosted compute. The graphforge.sh Hub distributes repositories; execution stays local. 13 | Yes SurrealDB Cloud: single-node Start instances to multi-node Scale clusters. 19 |
| Replication | Not planned HA replication is out of scope. 9 | Paid edition Replication/fault tolerance come from distributed storage on Cloud Scale or Enterprise. 19 |
What will my team write?
GraphForge: openCypher from Python, Node.js, Rust. SurrealDB: SurrealQL, ISO GQL (since 3.2; default-on since 3.3), GraphQL.
SurrealDB has ten SDKs and embeds from Rust, .NET, Go, Python, and JavaScript, plus an opt-in Postgres wire protocol that carries SurrealQL or GQL. GraphForge has three bindings (Python, Node.js, Rust) and standard openCypher, so SurrealQL code does not port directly.
- Query languages
- GraphForge openCypher
- SurrealDB SurrealQL, ISO GQL (since 3.2; default-on since 3.3), GraphQL
- SDKs and bindings
- GraphForge Python, Node.js, Rust
- SurrealDB Rust, JavaScript, Python, Go, .NET, Java, Kotlin, PHP, Swift, Mojo
- Written in
- GraphForge Rust
- SurrealDB Rust
- Surfaces
- GraphForge Python, Node.js, Rust crates, CLI, VS Code extension, Notebooks, Agent skills
What are the trade-offs?
SurrealDB has no algorithm library and a BSL license. GraphForge has one data model and no browser build.
A fair assessment cuts both ways.
GraphForge
- No document, geospatial, or time-series model.
- No WebAssembly or browser build.
- No server, cluster, or managed cloud.
SurrealDB
- The BSL licence is not an open-source licence and prohibits providing SurrealDB as a database service to third parties until the change date.17
- There is no built-in centrality, community-detection or similarity algorithm library; +shortest returns only one path when several tie.21
- Time-travel queries require versioning enabled on RocksDB or beta SurrealKV; the in-memory SurrealMX engine refuses versioned=true from 3.3.0.22
- Self-hosted Community TiKV is positioned only for local multi-node experimentation; production HA clusters need Cloud Scale or Enterprise.20
- Table views update only on writes to the FROM table, so changes to joined records and OPTION IMPORT writes do not refresh them.24
How would I migrate?
Load SurrealDB data through Arrow, pandas, or Polars. SurrealQL queries are rewritten as Cypher.
Into GraphForge
- GraphForge 0.5.x loads data through Arrow, pandas, and Polars bulk calls or Cypher scripts; it has no native CSV loader.15
- Algorithms are API verbs (rank, cluster, paths, analyze, similar), not Cypher CALL procedures, so procedure calls become verb calls.7
- GraphForge passes the full openCypher 2024.3 TCK, so standard openCypher queries are the portable part of a migration.10
Moving from SurrealDB
- GQL runs over existing tables and RELATE edges but differs from openCypher: -- is a comment, inequality is <>, and quantifiers follow the edge.26
- SurrealQL changed across majors: SEARCH ANALYZER became FULLTEXT ANALYZER in 3.0.0 and the DEFER index clause is rejected on 3.x.23
- A 3.2 datastore upgrades in place to 3.3 but a 3.2 binary misreads migrated data, so export before upgrading if rollback may be needed.18
What else do people ask about SurrealDB?
Does SurrealDB have graph algorithms? SurrealDB has path-finding only, runs embedded and in browsers, and is BSL 1.1 until 2030.
Does SurrealDB have graph algorithms?
It has path-finding: the +shortest, +path, and +collect modifiers and GQL SHORTEST patterns. Its documentation shows no centrality, community-detection, or similarity library.
Is SurrealDB open source?
SurrealDB 3.x uses the Business Source License 1.1, which forbids offering it as a database service and converts to Apache-2.0 on 1 January 2030. GraphForge is Apache-2.0.
Can SurrealDB run embedded like GraphForge?
Yes. SurrealDB embeds in-process from several SDKs and in browsers through WebAssembly. GraphForge embeds from Python, Node.js, and Rust, but has no browser build.
How was this researched?
From 28 primary sources, accessed 2026-10-19, each cited where it is used.
Capabilities were checked against primary documentation, repositories, and release notes, accessed 2026-10-19. Each cell records two things: the capability’s state and the evidence behind it. Every GraphForge cell carries a release status. Shipped means stable 0.5.2; experimental means 0.6.0 release candidates only. GraphForge is pre-1.0 and makes no backward-compatibility guarantee yet.
- Yes documented natively
- Partial limited or caveated
- Extension official extension or module
- Paid edition paid or enterprise edition only
- Removed removed, with the version cited
- Not established not confirmed by our research; not a “no”
- Shipped in stable 0.5.2
- Partial shipped with documented limits
- Experimental in 0.6.0 release candidates only
- Planned on the roadmap
- Not offered outside the product's model
- Not planned explicitly out of scope
Benchmark classes
- Equivalent workloads. The same logical operation with the same correctness requirement on every engine.
- Native optimized workloads. Each engine's best supported implementation, including projections, indexes, extensions, and native algorithms.
- End-to-end application workloads. Time from raw data to a usable result: import, indexing, projection, computation, and output retrieval.
Sources · accessed 2026-10-19
- GraphForge repository README — GitHub
- GraphForge releases — GitHub
- GraphForge 0.5.0 changelog — GitHub
- Storage architecture — GraphForge Book
- Analytics integration: search and retrieval — GraphForge Guide
- ADR 0018: acknowledged durability and isolation — GitHub
- Algorithm verbs — GraphForge Book
- Cypher guide — GraphForge Guide
- openCypher compatibility — GraphForge Reference
- openCypher TCK compliance — GraphForge Reference
- Remote authority peer extension (#215) — GitHub
- ADR 0027: native runtime boundary — GitHub
- Release scope and roadmap — GraphForge Docs
- Keep research history — GraphForge Guide
- Graph construction — GraphForge Guide
- Knowledge public API v1 — GraphForge Book
- LICENSE (Business Source License 1.1) — surrealdb/surrealdb GitHub
- Release 3.3 — SurrealDB
- Architecture — SurrealDB Docs
- start command — SurrealDB CLI Docs
- Idioms (recursive paths) — SurrealQL Docs
- SELECT statement (VERSION clause) — SurrealQL Docs
- DEFINE INDEX statement — SurrealQL Docs
- DEFINE TABLE statement — SurrealQL Docs
- Embedding SurrealDB — SurrealDB Docs
- GQL overview — SurrealDB Docs
- Data models — SurrealDB Docs
- RELATE statement — SurrealQL Docs
Try GraphForge on your own graph.
Install the package, open a project, and run the same notebook we measure with.