> ## Documentation Index
> Fetch the complete documentation index at: https://docs.nexusapi.link/llms.txt
> Use this file to discover all available pages before exploring further.

# API basics: authentication, models, and protocols

> Confirm the Base URL, API key, model ID, and protocol before making a NexusAPI call.

This page covers the rules shared by every model call. Use the [Apifox API reference](https://nexusapi.apifox.cn/) for complete request and response schemas; this documentation explains **which protocol to choose and how to start safely**.

<Info>
  **Confirm capability first**

  Available models change with your API key group, model restrictions, and live service state. The whole-site Model Marketplace is not a promise that every model is callable by a particular key.
</Info>

## Three required values

| Value    | Where to get it                   | Purpose                                     |
| -------- | --------------------------------- | ------------------------------------------- |
| Base URL | This page                         | Sends SDK or client requests to NexusAPI    |
| API key  | Console → Token management        | Identifies the caller and accessible groups |
| Model ID | Model Marketplace or `/v1/models` | Selects the model for a request             |

For a standard OpenAI-style client, use:

```text theme={"system"}
https://nexusapi.link/v1
```

For manual HTTP calls, send the key in `Authorization`:

```http theme={"system"}
Authorization: Bearer YOUR_API_KEY
```

Never put a complete key in frontend code, screenshots, chat history, or a Git repository. Client URL rules vary: Claude Code uses the site address without `/v1`, while Codex CLI uses its Provider configuration. Follow the relevant [client tool guide](/en/tools/index).

## List models visible to this key

```bash theme={"system"}
curl 'https://nexusapi.link/v1/models' \
  -H 'Authorization: Bearer YOUR_API_KEY'
```

Copy a model ID from `data[].id`. This is the set **currently visible to that key**, not a full platform catalog. If a response includes `supported_endpoint_types`, it can help identify the protocol type shown for the model.

## Select the right protocol

| Goal                                        | Path                          | Use it for                                                                 |
| ------------------------------------------- | ----------------------------- | -------------------------------------------------------------------------- |
| OpenAI-compatible text, chat, and streaming | `POST /v1/chat/completions`   | Most custom apps, OpenAI SDKs, and generic clients                         |
| Native Anthropic / Claude Code              | `POST /v1/messages`           | Models labelled `anthropic`                                                |
| Codex Provider                              | `POST /v1/responses`          | Only with the [Codex CLI guide](/en/tools/codex-cli) and a confirmed model |
| Image generation                            | `POST /v1/images/generations` | Image-capable models visible to the current group                          |

The same model name does not mean every request format is accepted. Confirm the protocol in the [API and model capability matrix](/en/developer/capability-matrix), then read:

* [Chat Completions](/en/developer/openai-compatible)
* [Anthropic Messages](/en/developer/anthropic-messages)
* [Responses / Codex](/en/developer/responses)
* [Image model guide](/en/guide/image-generation)

## Minimal verification and retries

1. List models, then send one short text request or minimum image request.
2. Confirm the model ID, key group, and path before adding optional fields, streaming, or batch work.
3. Keep the error text, time, and Request ID on failure. Do not replay an unconfirmed request at high frequency.

For `401`, check the key first. For `404`, check a missing or repeated `/v1`. For `400`, check the model, protocol, and fields. See the [self-check guide](/en/faq/self-check) for more.