GPT API and MCP: how to connect the chat models to your project

GPT is a family of OpenAI-compatible chat models. NeuralSpace offers the following models: GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna, GPT-5 Mini, GPT-5.4 Mini. They all run through a single public API, and there is a ready MCP server for AI agents. The key is issued on the API keys page.

What GPT can do

The three GPT-5.6 tiers — Sol, Terra and Luna — differ in power and price: Sol for hard tasks, Terra as the balance, Luna for fast replies. GPT-5 Mini and GPT-5.4 Mini are compact models for high-volume requests. All of them understand and write text, work with code and hold conversations.

API endpoint

Send a request to POST /v1/chat/completions. Example: {"model": "gpt-5-6", "messages": [{"role": "user", "content": "Hello"}]}. Authorization uses an API key in the x-api-key header. Available model ids: gpt-5-6, gpt-5-6-balanced, gpt-5-6-fast, gpt-5-mini, gpt-5-4-mini.

MCP tools

The NeuralSpace MCP server is available at POST /mcp (Streamable HTTP). To work with GPT, an agent uses the tools: chat, list_models, get_balance. This lets Claude Code, Cursor and other MCP clients use the site models directly.

How to connect

Sign up on NeuralSpace, open the API keys page and create an nsk-… key. Then pass it in the request header or in your MCP client settings. Full documentation is on the the /v1/docs page.

FAQ

How does Sol differ from Luna?

Sol is the most powerful tier for hard tasks, Luna is fast and cheap for high-volume requests.

What format is the request?

OpenAI-compatible: POST /v1/chat/completions with a messages array.

Where do I get an API key?

On the API keys page in the dashboard.