# Zelinqa MCP

> Zelinqa MCP is a freemium MCP server for ai models and llm tooling. Let Zelinqa choose each next question so an assistant gathers information in fewer turns. It runs locally on your machine and is started by your AI client.

- **Kind:** MCP server
- **Pricing:** Freemium
- **Platforms:** CLI, API, self-hosted
- **Website:** https://docs.zelinqa.ai
- **Repository:** https://github.com/zelinqa/zelinqa-mcp
- **MCP Registry name:** io.github.Zelinqa/zelinqa-mcp
- **Categories:** AI models and LLM tooling, sales and CRM
- **Listing updated:** 2026-11-16

## About

In most assistants the model improvises what to ask next and often decides with gaps. Zelinqa moves that job to an engine. You describe in Zelinqa Studio what a conversation must find out: dimensions such as need, budget and timing, the questions that may be asked, the structured data to collect and a depth level. The MCP server then lets any host run the conversation.

The flow uses a few calls. zelinqa_start opens a conversation and returns the first question with choices, the model asks it in its own words, and zelinqa_next_question takes the person's reply as text or chosen labels and returns the next question with progress per dimension, until the engine stops. At the end the model records what actually happened. The model sees only question text, choice labels and a conversation name; session and decision identifiers stay inside the server.

Selection is deterministic over your published domain, so the same state yields the same question and each decision carries what it collects, progress and a stop reason, which helps audits. The vendor cites an internal test bench of more than 700 replayed conversations showing fewer tokens, and plans a public benchmark. It needs Python 3.11 with uv and a runtime API key; a free Developer edition account is enough.

## Features

- **Engine-chosen questions:** Selects the next question from your published domain.
- **Deterministic choices:** The same state always produces the same question.
- **Progress tracking:** Reports coverage per dimension and a stop reason.
- **Typed state:** Collects structured data for your application.
- **Model sees little:** Only question text, labels and a conversation name.
- **Fewer tokens:** The vendor reports savings in an internal test bench.

## Good for

- Qualify a sales lead with a short guided interview
- Run an intake or triage conversation consistently
- Collect a project brief with budget and timeline
- Audit why a conversation asked what it did

## Install

### Claude Code (Terminal)

```sh
claude mcp add zelinqa-mcp --env ZELINQA_API_KEY='<ZELINQA_API_KEY>' -- uvx zelinqa-mcp
```

### Claude Desktop (claude_desktop_config.json)

```json
{
  "mcpServers": {
    "zelinqa-mcp": {
      "command": "uvx",
      "args": [
        "zelinqa-mcp"
      ],
      "env": {
        "ZELINQA_API_KEY": "<ZELINQA_API_KEY>"
      }
    }
  }
}
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "zelinqa-mcp": {
      "command": "uvx",
      "args": [
        "zelinqa-mcp"
      ],
      "env": {
        "ZELINQA_API_KEY": "<ZELINQA_API_KEY>"
      }
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "zelinqa-mcp": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "zelinqa-mcp"
      ],
      "env": {
        "ZELINQA_API_KEY": "<ZELINQA_API_KEY>"
      }
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "zelinqa-mcp": {
      "command": "uvx",
      "args": [
        "zelinqa-mcp"
      ],
      "env": {
        "ZELINQA_API_KEY": "<ZELINQA_API_KEY>"
      }
    }
  }
}
```

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "zelinqa-mcp": {
      "command": "uvx",
      "args": [
        "zelinqa-mcp"
      ],
      "env": {
        "ZELINQA_API_KEY": "<ZELINQA_API_KEY>"
      }
    }
  }
}
```

## Tools

- `zelinqa_start`: Start a conversation and return the first question
- `zelinqa_next_question`: Pass the person's reply and get the next question with progress

## FAQ

### What do I need to start?

Python 3.11 with uv and a runtime API key for a published domain from Zelinqa Studio.

### Is there a free tier?

The README says a free Developer edition account is enough.

### Is the question choice generated by a model?

No. It is a selection over your published domain and is deterministic.

### What does the model see?

Only question text, choice labels and a conversation name.

### Is Zelinqa MCP free?

Zelinqa MCP has a free plan, with paid plans for more usage or features.

### How do I add Zelinqa MCP to Claude Code?

Run this command in your terminal: claude mcp add zelinqa-mcp --env ZELINQA_API_KEY='<ZELINQA_API_KEY>' -- uvx zelinqa-mcp

### What are the best alternatives to Zelinqa MCP?

The closest alternatives to Zelinqa MCP are Claude Code Ultimate Guide, Calven, CompanyData, MisarReach MCP Server and NeuralVerge.

## Alternatives

- [Claude Code Ultimate Guide](https://ai-powered.com/mcp/claude-code-ultimate-guide): Claude Code Ultimate Guide is an open-source MCP server for ai models and llm tooling.
- [Calven](https://ai-powered.com/mcp/calven): Calven is an MCP server for marketing and seo.
- [CompanyData](https://ai-powered.com/mcp/companydata): CompanyData is a paid MCP server for analytics and data.
- [MisarReach MCP Server](https://ai-powered.com/mcp/misarreach-mcp-server): MisarReach MCP Server is a freemium MCP server for sales and crm.
- [NeuralVerge](https://ai-powered.com/mcp/neuralverge): NeuralVerge is an MCP server for web search and scraping.
- [ZUUZ](https://ai-powered.com/mcp/zuuz): ZUUZ is an MCP server for sales and crm.
- [Fuse](https://ai-powered.com/mcp/fuse-2): Fuse is a freemium MCP server for sales and crm.
- [signafAIde](https://ai-powered.com/mcp/signafaide): signafAIde is a freemium MCP server for security.

---
Source: https://ai-powered.com/mcp/zelinqa-mcp
