curl -X POST https://synapse.by-kit.com/api/v1/record \
  -H "Authorization: Bearer sk_syn_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"agent_id": "my-agent", "type": "fact", "title": "User pref", "body": "User prefers dark mode"}'

Or skip the steps entirely: hand this docs link to your coding agent and tell it to implement Synapse. It reads the instructions and configures itself — you just paste in the API key when it asks.

How Synapse compares

Feature	Other solutions	Synapse
LLM call per write	1 call per write (~$0.003)	0 calls ($0.00)
Advanced features	Top tier only ($249+/mo)	All tiers
Framework lock-in	Vendor SDK required	Any HTTP client
Model portability	Locked to one vendor	Works with any LLM
Deploy path	SDK install + config	One API key, any HTTP client
Classifier accuracy	~49–70% typical	Domain-trained

Benchmarks based on independent LongMemEval testing across leading memory providers.

Your memory. Any model.

Switch from Claude to ChatGPT to Gemini. Your memories come with you. Synapse isn't tied to any LLM — it's infrastructure, not a feature.

Claude ChatGPT Gemini LangChain AutoGen

How it works

You're probably paying your LLM to remember things it already knew. Every context reload, every repeated summary — that's tokens. Synapse fixes that.

Write

Call POST /v1/record. Body stored in Postgres, classified by a self-hosted Gemma model on infrastructure we control. No third-party LLM call. No token cost.