![\"Gnosem\"](\"https://raw.githubusercontent.com/gnosem/gnosem/main/brand/gnosem-lockup-inverse.svg\")
\n \n \n
\n Cross-vendor AI memory over MCP.
\n One memory store. Every model. Claude, ChatGPT, Gemini, Kimi, Cursor, Windsurf — anything that speaks MCP.\n
\n gnosem.dev\n ·\n pricing\n ·\n llms.txt\n
\n\n---\n\n## What it is\n\nGnosem is a hosted [Model Context Protocol](https://modelcontextprotocol.io) server that gives each user one persistent memory store — readable and writeable from every MCP-capable AI client. Vendor-agnostic by design: a fact saved from Claude is visible in ChatGPT, Cursor, Windsurf, Zed, Kimi, and any other MCP client (or plain HTTP tool call) using the same API key.\n\nThe problem it solves: every AI vendor has its own proprietary memory. OpenAI's ChatGPT memory doesn't work in Claude. Anthropic's projects don't cross into GPT. Multi-model users retell context in every session. Gnosem is the neutral layer between vendors.\n\n## Tools exposed\n\n- `memory_write` — save a fact, preference, decision, or note. `content`, `tags?`, `written_by?` (provenance), `session_id?`, `no_optimize?`\n- `memory_search` — semantic search across your memories. `query`, `k?`, `raw?`\n- `memory_list` — list recent memories in reverse-chronological order. `limit?`, `cursor?`, `raw?`\n- `memory_forget` — soft-delete a memory by id\n- `memory_supersede` — replace a stale memory with a corrected one\n\n## AI-optimized storage\n\nLong memories (>400 chars) are automatically compressed on write to a structured-facts form (Workers AI, llama-3.1-8b-instruct-fast) with a strict prompt: `TOPIC=..., PROJECT=..., DECISION=..., STACK=..., PROBLEM=...`. The reading LLM ingests the compact form by default — fewer tokens, same meaning. The raw prose is preserved and returned as `content_raw`; pass `raw:true` to invert. Pass `no_optimize:true` on write to skip compression entirely.\n\nCompression is guarded — if the model output isn't actually shorter, gnosem falls back to storing raw only. Fail-open: any AI error still saves the memory.\n\n## Quickstart\n\n```bash\nnpx gnosem-install\n```\n\nAuto-detects and configures every MCP-capable client on your machine (Claude Desktop, Claude Code, Cursor, Windsurf, Zed). Prompts for your API key and merges the `gnosem` entry alongside your existing servers. Restart the affected clients — the five memory tools appear immediately.\n\nGet an API key at [gnosem.dev](https://gnosem.dev), or via the API:\n```bash\ncurl -sX POST https://gnosem.dev/signup \\\n -H \"Content-Type: application/json\" \\\n -d '{\"email\":\"you@example.com\"}'\n```\nThe key is displayed once. Save it.\n\n### Manual config (any MCP client)\n\n```json\n{\n \"mcpServers\": {\n \"gnosem\": {\n \"url\": \"https://gnosem.dev/mcp\",\n \"headers\": { \"Authorization\": \"Bearer gn_your_api_key\" }\n }\n }\n}\n```\n\n## Quickstart — any HTTP client\n\n```bash\n# Write\ncurl -sX POST https://gnosem.dev/mcp \\\n -H \"Authorization: Bearer gn_your_key\" -H \"Content-Type: application/json\" \\\n -d '{\"jsonrpc\":\"2.0\",\"id\":1,\"method\":\"tools/call\",\"params\":{\"name\":\"memory_write\",\"arguments\":{\"content\":\"I prefer Postgres for greenfield work.\",\"written_by\":\"claude-code\",\"tags\":[\"preference\"]}}}'\n\n# Search\ncurl -sX POST https://gnosem.dev/mcp \\\n -H \"Authorization: Bearer gn_your_key\" -H \"Content-Type: application/json\" \\\n -d '{\"jsonrpc\":\"2.0\",\"id\":2,\"method\":\"tools/call\",\"params\":{\"name\":\"memory_search\",\"arguments\":{\"query\":\"database preference\",\"k\":5}}}'\n```\n\n## Pricing\n\n| Plan | Price | Memories | API keys | Storage |\n|---|---|---|---|---|\n| Free | $0 | 200 | 1 | shared |\n| Pro | $9/mo or $90/yr | Unlimited | Unlimited | 1 GB |\n\nUpgrade at [gnosem.dev/upgrade](https://gnosem.dev/upgrade).\n\n## Data model\n\nEach memory row: `id` (uuid), `content` (raw, ≤8000 chars), `content_optimized` (structured facts, nullable), `tags` (json array), `written_by` (provenance — which model wrote it), `session_id` (opaque grouping), `created_at`, plus a semantic embedding ([BGE-base-en-v1.5](https://huggingface.co/BAAI/bge-base-en-v1.5), 768-dim) stored in Vectorize.\n\nEvery memory tracks provenance so you can see which model contributed which fact. Memories are per-user isolated at both the D1 metadata layer (`user_id` filter on every query) and the Vectorize layer (metadata index on `user_id`).\n\nCorrection chains via `memory_supersede`: the old row is marked superseded (excluded from reads) and the new row references it. Soft-delete via `memory_forget` (excluded from reads and removed from the vector index).\n\n## Infrastructure\n\n- **[Cloudflare Workers](https://workers.cloudflare.com/)** — edge-hosted MCP server, no cold starts\n- **[D1](https://developers.cloudflare.com/d1/)** — metadata + auth\n- **[Vectorize](https://developers.cloudflare.com/vectorize/)** — 768-dim embedding storage + cosine ANN search, per-user metadata filter\n- **[Workers AI](https://developers.cloudflare.com/workers-ai/)** — embedding generation (`@cf/baai/bge-base-en-v1.5`) and content optimization (`@cf/meta/llama-3.1-8b-instruct-fast`)\n- **[Stripe](https://stripe.com)** — subscription billing\n\nLatency: sub-100ms for read, ~200-800ms for write (dominated by embedding + optimization inference).\n\n## AI discovery\n\nThe service is designed to be discoverable and citeable by LLMs:\n\n- [`/llms.txt`](https://gnosem.dev/llms.txt) — LLM-optimized site summary per [llmstxt.org](https://llmstxt.org)\n- [`/robots.txt`](https://gnosem.dev/robots.txt) — explicitly welcomes GPTBot, ClaudeBot, PerplexityBot, Google-Extended, and every major AI crawler\n- [`/sitemap.xml`](https://gnosem.dev/sitemap.xml) — standard sitemap\n- Schema.org JSON-LD (`SoftwareApplication` + `FAQPage`) embedded in the landing page\n\n## Self-hosting\n\nThe Worker source in this repo can be deployed to your own Cloudflare account. You'll need:\n\n1. A Cloudflare account with Workers, D1, Vectorize, and Workers AI enabled\n2. `npx wrangler d1 create