iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Yes—you can connect LlamaIndex’s VectorStoreIndex and PropertyGraphIndex to one SynapCores service using two separate Python packages. In the integration example, LlamaIndex generates embeddings with the model configured in its settings; SynapCores stores them, so this path does not require installing an embedding model in the database engine.
How the two integrations fit together
The SynapCores team describes two adapters: llama-index-vector-stores-synapcores maps SynapCores to LlamaIndex’s vector-store abstraction, while llama-index-graph-stores-synapcores maps it to the property-graph-store abstraction. The packages are separate, even when both adapters connect to the same database service.
The distinction is about retrieval needs. A vector index retrieves chunks by embedding similarity and can apply metadata filters. A property graph index represents entities and their relationships, enabling relationship-aware expansion and structured graph queries. LlamaIndex’s general pattern is to pass a PropertyGraphStore to PropertyGraphIndex for insertion and querying; that framework behavior is distinct from claims about the SynapCores adapter. LlamaIndex Property Graph Index guide
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Install the package for the retrieval pattern you need
Install the vector-store package for vector retrieval, the graph-store package for property-graph retrieval, or both if your application uses both index types. The vendor’s June 17, 2026 article describes the packages at v0.1.0; check the package’s current release information before relying on version-specific behavior.
#1 Best Overall
pip install llama-index-vector-stores-synapcores
pip install llama-index-graph-stores-synapcoresThe graph package’s published quickstart also installs llama-index alongside the graph adapter and starts a SynapCores Community container. Consult the package instructions for the current container command and prerequisites: llama-index-graph-stores-synapcores on PyPI.
Connect each adapter to SynapCores
The vendor’s example points both adapters at http://localhost:8080. Use the URI reachable from your application environment; for example, a containerized application may need a service hostname instead of localhost. The vector adapter takes a table name and embedding dimension, while the graph adapter takes a graph name and embedding dimension.
from llama_index.core import StorageContext, VectorStoreIndex, PropertyGraphIndex, Settings
from llama_index.core import SimpleDirectoryReader
from llama_index.vector_stores.synapcores import SynapCoresVectorStore
from llama_index.graph_stores.synapcores import SynapCoresPropertyGraphStore
# Configure Settings.embed_model before indexing, as appropriate for your setup.
documents = SimpleDirectoryReader("./data").load_data()
vector_store = SynapCoresVectorStore(
uri="http://localhost:8080",
table_name="documents",
embedding_dim=1536,
)
storage_context = StorageContext.from_defaults(vector_store=vector_store)
vector_index = VectorStoreIndex.from_documents(
documents,
storage_context=storage_context,
)
graph_store = SynapCoresPropertyGraphStore(
uri="http://localhost:8080",
graph_name="knowledge",
embedding_dim=1536,
)
graph_index = PropertyGraphIndex.from_documents(
documents,
property_graph_store=graph_store,
)This illustrates the vendor’s setup pattern, not the only valid LlamaIndex configuration. Confirm the installed adapter’s import paths and constructor arguments in its current documentation. The embedding dimension must match the vectors produced by the configured embedding model and the store’s schema; the example value above is illustrative, not a recommended dimension.
Recommended Free Tools
Do you need to install a model in SynapCores?
No, not for the integration path described by the vendor. The embedding model configured in LlamaIndex computes document embeddings, which the adapter writes to a SynapCores VECTOR(N) column. The SynapCores team’s June 17, 2026 article says the integration does not require downloading an engine-side model. You still need to configure an embedding model that your LlamaIndex application can use; the article names OpenAI, Hugging Face, and Cohere as examples, not endorsements or requirements.
What each adapter claims to support
The capabilities below are described by the SynapCores integration author; they are not independent audits of adapter behavior.
| Adapter | Vendor-described operations and features | Relevant LlamaIndex index |
|---|---|---|
| Vector store | add(nodes), delete(ref_doc_id), query(VectorStoreQuery), optional delete_nodes, clear, and get_nodes; upserts; HNSW indexing on a VECTOR(N) column; cosine, Euclidean, and dot-product distance; metadata filters; asynchronous wrappers; and reuse of existing tables through VectorStoreIndex.from_vector_store. |
VectorStoreIndex |
| Property graph store | Entity and chunk upserts, typed relations, filtered gets, depth-bounded relation-map expansion, structured Cypher queries, vector search over chunk embeddings, and an asynchronous surface. The author says supports_structured_queries and supports_vector_queries are true. |
PropertyGraphIndex |
SynapCores’ product documentation describes its version 1.5 property-graph engine and Cypher surface, including MATCH, WHERE, RETURN, ORDER BY, LIMIT, and MERGE. That describes the database product, not independent validation of the Python adapter. SynapCores features
Choose one database service or separate backends
Using one SynapCores service for both adapters can simplify the topology when the application needs both vector and relationship-aware retrieval and the supported features fit its queries. It does not mean the two indexes are interchangeable: choose the vector index for similarity-oriented retrieval, and the property graph index when entities, relations, or graph expansion matter.
A separate vector and graph backend may be more appropriate if your required query behavior, operational constraints, or existing architecture calls for different systems. The available sources establish no independent benchmark showing that one arrangement is faster, cheaper, or more scalable. Measure your own representative workload and assess feature fit, service operations, reliability, and performance before choosing.
Check version maturity before depending on roadmap items
The June 17, 2026 vendor article described the packages as v0.1.0. It said asynchronous methods in that release used asyncio.to_thread around the synchronous SDK; a native httpx async path was listed for v0.3.0, and a hybrid retriever for v0.2.0. Those are dated roadmap statements, not confirmation that the milestones shipped. Check current releases and documentation if either capability is important. The article also described official LlamaIndex integration listing as a future goal.
The vendor reported 48/48 tests against a stock synapcores/community:latest container in 2026, with an empty models directory and no model pulls. This is the vendor’s reported result for its test pass—not an independent reliability study or a performance benchmark.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
Free tools Windows power users keep installed
One-click scans. No signup required.
