{"title":"General FAQ","slug":"general","category":"faq","summary":"Common questions about Synapse — what it is, how it works, who it's for.","audience":["human","llm"],"tags":["faq","general","intro"],"difficulty":"beginner","updated":"2026-06-27","word_count":543,"read_minutes":3,"lang":"en","translated":true,"requested_lang":"en","content_markdown":"\n# General FAQ\n\nCommon questions about Synapse.\n\n## What is Synapse?\n\nSynapse is a persistent memory API for LLM agents. It gives your AI assistant\na permanent, queryable brain that survives across sessions. Instead of\nforgetting everything when the chat closes, the LLM stores and retrieves\nmemories via a simple HTTP API.\n\n**Learn more:** [What is Synapse?](/docs/getting-started/what-is-synapse)\n\n## Who is Synapse for?\n\n- **LLM agent developers** who need persistent state\n- **Power users** running local LLMs with custom agents\n- **Teams** building AI assistants with shared memory\n- **Automation engineers** chaining LLM calls across sessions\n\n## Is Synapse free?\n\nSynapse is hosted at and free for personal use.\nFor self-hosting, see the [repo](https://gitlab.com/schaefer-services/synapse).\n\n## How is Synapse different from ChatGPT Memory?\n\n| Feature | ChatGPT Memory | Synapse |\n|---------|---------------|---------|\n| Storage | OpenAI servers | Your server |\n| API access | No | Yes (REST + MCP) |\n| Multi-tenant | No | Yes (minds) |\n| Custom categories | No | Yes (8 categories) |\n| Full-text search | Limited | FTS5 + semantic |\n| Self-hostable | No | Yes |\n\n## What LLMs work with Synapse?\n\nAny LLM that can make HTTP calls or use MCP:\n\n- Claude (Anthropic) — via MCP or direct API\n- GPT-4 / GPT-3.5 (OpenAI) — via function calling + API\n- Gemini (Google) — via function calling + API\n- Llama / Mistral (local) — via custom integration\n- Any LLM via the Synapse MCP server\n\n## Do I need to host Synapse myself?\n\nNo. The hosted version at is available for\npublic use. Self-hosting is optional (for privacy, customization, or\nair-gapped environments).\n\n## How much data can I store?\n\nThere are no hard limits on the hosted version. Typical usage:\n\n- 100-1000 memories per mind\n- 1-10 KB per memory (content field)\n- Total: ~10 MB per mind\n\nFor larger scale, self-host.\n\n## Is my data private?\n\nYes. Each mind is isolated (see [Multi-Tenancy](/docs/concepts/multi-tenancy)).\nOther users cannot see your data. The hosting provider (Schäfer Services)\nhas access but does not view user data.\n\nFor maximum privacy, self-host.\n\n## Can I export my data?\n\nYes. Use `GET /memory/mind-export` to export all memories as JSON. See\n[Backup & Restore](/docs/guides/backup-restore).\n\n## What happens if I lose my Mind Key?\n\nMind Keys are shown only once at creation. If lost:\n\n1. Login with your email/password to get a JWT\n2. Create a new mind via `POST /minds`\n3. Save the new Mind Key\n4. (Optional) Delete the old mind via `DELETE /minds/:id`\n\nYou cannot recover memories from a mind whose key is lost.\n\n## Can multiple LLM agents share a mind?\n\nYes. All agents using the same Mind Key share that mind's data. For\ncoordinated multi-agent work, see [Multi-Agent Coordination](/docs/guides/multi-agent-coordination).\n\n## Does Synapse work offline?\n\nNo, Synapse requires an internet connection to the server (either hosted or\nself-hosted). For offline LLMs, run Synapse locally.\n\n## How fast is memory search?\n\n- FTS5 keyword search: < 10ms for 1000 memories\n- Semantic search: 50-100ms for 1000 memories\n- Full recall: < 50ms for 1000 memories\n\nSee [FTS5 Search](/docs/concepts/fts5-search) for details.\n\n## Can I use Synapse without an LLM?\n\nYes. Synapse is a generic memory API. You can use it from any application\nthat benefits from persistent, searchable key-value storage with categories\nand tags.\n\n## Next Steps\n\n- [What is Synapse?](/docs/getting-started/what-is-synapse)\n- [Quick Start](/docs/getting-started/quick-start)\n- [API FAQ](/docs/faq/api-faq)\n","content_html":"

General FAQ

\n

Common questions about Synapse.

\n

What is Synapse?

\n

Synapse is a persistent memory API for LLM agents. It gives your AI assistant\na permanent, queryable brain that survives across sessions. Instead of\nforgetting everything when the chat closes, the LLM stores and retrieves\nmemories via a simple HTTP API.

\n

Learn more: What is Synapse?

\n

Who is Synapse for?

\n\n

Is Synapse free?

\n

Synapse is hosted at https://synapse.schaefer.zone and free for personal use.\nFor self-hosting, see the repo.

\n

How is Synapse different from ChatGPT Memory?

\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
FeatureChatGPT MemorySynapse
StorageOpenAI serversYour server
API accessNoYes (REST + MCP)
Multi-tenantNoYes (minds)
Custom categoriesNoYes (8 categories)
Full-text searchLimitedFTS5 + semantic
Self-hostableNoYes
\n

What LLMs work with Synapse?

\n

Any LLM that can make HTTP calls or use MCP:

\n\n

Do I need to host Synapse myself?

\n

No. The hosted version at https://synapse.schaefer.zone is available for\npublic use. Self-hosting is optional (for privacy, customization, or\nair-gapped environments).

\n

How much data can I store?

\n

There are no hard limits on the hosted version. Typical usage:

\n\n

For larger scale, self-host.

\n

Is my data private?

\n

Yes. Each mind is isolated (see Multi-Tenancy).\nOther users cannot see your data. The hosting provider (Schäfer Services)\nhas access but does not view user data.

\n

For maximum privacy, self-host.

\n

Can I export my data?

\n

Yes. Use GET /memory/mind-export to export all memories as JSON. See\nBackup & Restore.

\n

What happens if I lose my Mind Key?

\n

Mind Keys are shown only once at creation. If lost:

\n
    \n
  1. Login with your email/password to get a JWT
    2. \n
  2. Create a new mind via POST /minds
    3. \n
  3. Save the new Mind Key
    4. \n
  4. (Optional) Delete the old mind via DELETE /minds/:id
    5. \n
\n

You cannot recover memories from a mind whose key is lost.

\n

Can multiple LLM agents share a mind?

\n

Yes. All agents using the same Mind Key share that mind's data. For\ncoordinated multi-agent work, see Multi-Agent Coordination.

\n

Does Synapse work offline?

\n

No, Synapse requires an internet connection to the server (either hosted or\nself-hosted). For offline LLMs, run Synapse locally.

\n

How fast is memory search?

\n\n

See FTS5 Search for details.

\n

Can I use Synapse without an LLM?

\n

Yes. Synapse is a generic memory API. You can use it from any application\nthat benefits from persistent, searchable key-value storage with categories\nand tags.

\n

Next Steps

\n\n","urls":{"html":"/docs/faq/general","text":"/docs/faq/general?format=text","json":"/docs/faq/general?format=json","llm":"/docs/faq/general?format=llm"},"translations_available":["en","zh","hi","es","fr","ar","pt","ru","ja","de","it","ko","nl","pl","tr","sv","vi","th","id","uk"]}