Skip to content

SynapCores / Community Edition

Download SynapCores.
Build on your own terms.

Free. Single binary. No signup, no credit card, no waitlist.

Choose your install method below.

Currently shippingv2.5.1-ce·published Nov 14, 2026

01 / Run in your terminal

curl -fsSL https://get.synapcores.com | SYNAPCORES_REF=download sh

Runs AI models locally: the installer offers a chat model sized to your machine's RAM (Qwen 2.5 7B is about 4.7 GB).

Windows: two commands

Set up Ubuntu on WSL2 once, then run the installer inside it. Microsoft's WSL guide

  1. Step 1 · PowerShell, as administrator

    wsl --install -d Ubuntu

    First time with WSL? Restart Windows when it finishes. Then open Ubuntu from the Start menu and create your Linux username and password.

  2. Step 2 · Ubuntu terminal

    curl -fsSL https://get.synapcores.com | SYNAPCORES_REF=download-windows sh

One-liner bootstrap. Auto-detects your OS and architecture (Linux x86_64 / aarch64, macOS Apple Silicon) and your GPU (CUDA-enabled hosts get the CUDA build automatically). Drops the binary, systemd unit (Linux) or launchd plist (macOS) and config template, then sets up your local chat model and the all-minilm embedding model.On Windows: run this inside a WSL2 (Ubuntu) terminal — it's the recommended path on Windows.Next: the quickstart goes from the first-boot login to your first SQL and vector query.

Need an offline install?

Both options above pull artifacts that are also available as direct downloads on the v2.5.1-ce GitHub Release page (with SHA-256 checksums) for air-gapped or locked-down hosts.

What's in Community Edition

Graph + Vector + LLM in one engine

Cypher, vector similarity, and LLM scoring all in a single query. No glue code, no second database.

255 ready-to-run recipes

Hello-world to GraphRAG to anti-money-laundering layering — searchable in the bundled web UI.

Local LLM and embeddings

Downloaded on install or first start: a local chat model sized to your memory (Qwen 2.5 7B with 8 GB+ to spare) and the all-minilm embedding model. Inference runs in-process; no cloud LLM calls until you add your own OpenAI / Anthropic key.

Multi-provider vision

Configure OpenAI, Anthropic Claude, or Ollama LLaVA from the Settings UI to caption images in your filesystem collections.

OCR + transcription bundled

Tesseract OCR for screenshots / scanned PDFs. Whisper transcription for audio + video (drop the model file).

Single-host scale

Designed for one VM / laptop. Multi-host clustering is the EE upgrade.

After install

The installer — or docker logs synapcores on the Docker path — prints the first-boot admin credentials (username, password, API key) and the URL of the bundled web UI (typically http://localhost:8080). From there:

Try a recipe: copy, paste, run

255 ready-to-run recipes ship inside the binary. Pick one, copy the SQL or Cypher into the bundled web UI, watch it run.

graph
5 min
Hello GraphRAG
advanced

Combine vector similarity, graph traversal, and LLM scoring in a single Cypher query.

graph
12 min
Fraud Rings ML Misses
intermediate

Score every transaction with ANOMALY_SCORE() in SQL, then find the ring it misses with one graph traversal.

ml
5 min
Customer Churn ML
beginner

AutoML training and inference in one SQL statement — no Python pipeline.

graph
6 min
Knowledge Graph from Earnings Call
beginner

Send a CFO call transcript to /v2/graph/extract. Query the extracted entities seconds later.

graph
10 min
Drug Repurposing
advanced

Walk drug-target-disease chains, then use SIMILAR_TO on mechanism embeddings.

graph
6 min
Semantic Patient Cohorts
beginner

Find clinical lookalikes by symptom embedding in real time.

Need clustering, RBAC, or paid support?

Enterprise Edition adds Raft clustering, CDC inbound, fine-grained RBAC, SSO, and encryption-at-rest. Enterprise Support is available as a paid SLA-backed contract for CE deployments — buy support without buying EE.

Contact sales

5-email course · free

Ship an agent with persistent memory in 5 emails

A 5-email course where you build a real agent with persistent memory — one runnable SQL recipe per email, no LangChain, no Pinecone, no Redis. Lesson 1 hits your inbox the second you sign up. After that: a new recipe every Monday. No spam.

Free forever. We'll never share your email.