---
title: "Nex API, Claude with Memory & Tools | StudioMeyer"
description: "Claude API with persistent memory, sessions, and a deeply integrated MCP toolset. One API key, REST endpoint, done. Customer data on German servers."
canonical: "https://studiomeyer.io/en/api"
language: en
last_updated: 2026-09-12
image: "https://studiomeyer.io/en/api/opengraph-image?bc60ae5b69b82e6d"
markdown_versions: ["https://studiomeyer.io/de/api.md", "https://studiomeyer.io/en/api.md", "https://studiomeyer.io/es/api.md"]
publisher: "StudioMeyer, https://studiomeyer.io (llms.txt: https://studiomeyer.io/llms.txt)"
---

StudioMeyer Memory API

# Memory for your AI. As an API.

A backend that remembers. Store facts, search semantically, build knowledge graphs, import ChatGPT, Claude and Gemini history. Your LLM provider stays your choice, we only ship the memory.

[Get API Key](https://studiomeyer.io/en/services/memory)[Read Docs](https://memory.studiomeyer.io/api)

50+

MCP Tools

10+

REST Endpoints

EU

Hosting

curl, memory search

```
curl -X POST https://memory.studiomeyer.io/api/search \
  -H "Authorization: Bearer sm_live_xxx" \
  -d '{"query": "what do we know about Q1 sales"}'
```

Response, 200 OK

```
{
  "success": true,
  "count": 3,
  "results": [
    { "type": "learning", "content": "Q1 revenue up 34% YoY...", "rank": 2.4 },
    { "type": "decision", "content": "Q1 pricing policy...", ... }
  ]
}
```

What this is

## Long-term memory as a backend service

We host the memory backend for your AI agents. Semantic search, knowledge graph, confidence decay, contradiction detection, multi-tenant isolation, all addressable via REST and MCP. You POST facts, we index with pgvector + trigram + full-text. You query, we return relevant hits. Your LLM does the rest. Ideal for Claude/GPT agents, chatbots with long-term context, internal knowledge tools and anything where memory matters.

Why memory as a service

## What you save by not building it yourself

### No RAG pipeline

No chunking. No embeddings. No vector DB to host. No hybrid-search magic. You POST JSON, we store, index, and return relevant hits.

### Knowledge graph included

Entities, observations, relations. Bi-temporal model. Automatic dedup via gatekeeper. Contradiction detection. All without a Neo4j setup.

### Import from ChatGPT/Claude/Gemini

POST your export files, we parse, structure, and write to your memory. ChatGPT JSON, Claude Projects, Gemini History, all supported.

REST Endpoints

## Memory over HTTP

All endpoints on memory.studiomeyer.io. Bearer token in Authorization header. Plus MCP endpoint at /mcp for Claude Desktop, Cursor, Codex.

POST`/api/learn`Store a fact (with category + tags)

POST`/api/search`Unified search: BM25 + semantic + graph

POST`/api/decide`Log a decision (title + reasoning)

POST`/api/entity`Create entity + observations

POST`/api/import`Import ChatGPT/Claude/Gemini history

POST`/api/backfill`Backfill embeddings after bulk ingest

GET`/api/export`Export everything as Markdown

GET`/api/account`Account info, plan, limits

Pricing

## Free, Pro and Team

Free plan, no card. Cancel anytime. Memory reads unlimited in your plan.

Free

0 EUR

- 1,000 learnings
- 100 entities
- 1 API key
- Single agent

[Choose plan](https://studiomeyer.io/en/services/memory)

Pro

9 EUR/mo

- 25,000 learnings
- 1,000 entities
- 3 API keys
- Multi-agent

[Choose plan](https://studiomeyer.io/en/services/memory)

Team

19 EUR/mo

- Unlimited learnings
- Unlimited entities
- 20 API keys
- Team sharing

[Choose plan](https://studiomeyer.io/en/services/memory)

Use Cases

## What teams use it for

### Claude/GPT agent with long-term context

Your agent forgets nothing. Customer data, project history, earlier decisions, all in memory, available every session. Via MCP in Claude Desktop/Cursor or via REST in your agent framework.

### Internal knowledge assistant

Company knowledge once via /api/import or /api/learn, then query via /api/search. Employees get consistent answers from the same pool. Tenant isolation per team or per department.

### Chatbot with CRM awareness

Store customer profiles, history, preferences. On every conversation the bot pulls context via /api/search. Feels like a support rep who knows the customer, because it does.

### Multi-agent shared memory

Multiple agents, one shared memory. Research agent stores findings, writer agent reads them, reviewer agent adds comments. All tenant-isolated, bi-temporally versioned.

Integration

## Integrated in 5 minutes

One HTTP request. Any language, any framework. No SDK required. Or: drop the MCP endpoint into Claude Desktop, done.

JavaScript, store + search

```
// 1. Store a fact
await fetch('https://memory.studiomeyer.io/api/learn', {
  method: 'POST',
  headers: { 'Authorization': 'Bearer sm_live_xxx', 'Content-Type': 'application/json' },
  body: JSON.stringify({
    category: 'insight',
    content: 'Acme Corp prefers video calls over phone',
    tags: ['acme', 'preferences']
  })
});

// 2. Search later
const res = await fetch('https://memory.studiomeyer.io/api/search', {
  method: 'POST', headers,
  body: JSON.stringify({ query: 'how does Acme prefer to meet?' })
});
```

FAQ

## Frequently Asked Questions

## Ready?

Free plan, no card. One email. One key. Memory live.

[Get API Key](https://studiomeyer.io/en/services/memory)[Read Docs](https://memory.studiomeyer.io/api)

## Sitemap

Every page of this site as Markdown: [sitemap](https://studiomeyer.io/sitemap.md). Curated entry point: [llms.txt](https://studiomeyer.io/llms.txt).