Source: https://wiki.niceai.dev/en/

# NiceAI API Quick Start

NiceAI is an AI model API relay. One API key lets you call Claude, GPT and Gemini models through APIs that follow the official OpenAI, Anthropic and Gemini formats. Change the base URL and API key in your tool or SDK, and your existing code keeps working.

The full documentation is in Chinese: [快速开始](/guide/quick-start). This page summarises what you need to get started.

## Base URLs

| Protocol | Base URL | Auth header |
| --- | --- | --- |
| OpenAI compatible (Chat Completions / Responses) | `https://api.niceai.dev/v1` | `Authorization: Bearer <key>` |
| Anthropic Messages (Claude Code, Anthropic SDK) | `https://api.niceai.dev` (no `/v1`) | `x-api-key: <key>` |
| Gemini | `https://api.niceai.dev/v1beta` | `x-goog-api-key: <key>` |

Create an API key at [niceai.dev/keys](https://niceai.dev/keys) and pick the group that matches your tool: **Claude** for Claude Code and the Anthropic SDK, **OpenAI** for Codex and the OpenAI SDK, **Gemini** for Gemini CLI. Available models are listed in the [model plaza](https://niceai.dev/model-plaza).

## First request

**-group**

```bash [curl]
curl https://api.niceai.dev/v1/chat/completions \
  -H "Authorization: Bearer $NICEAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "<model name>", "messages": [{"role": "user", "content": "Hello"}]}'
```

```python [Python]
import os
from openai import OpenAI

client = OpenAI(base_url="https://api.niceai.dev/v1", api_key=os.environ["NICEAI_API_KEY"])
resp = client.chat.completions.create(model="<model name>", messages=[{"role": "user", "content": "Hello"}])
print(resp.choices[0].message.content)
```

```js [Node.js]
import OpenAI from 'openai'

const client = new OpenAI({ baseURL: 'https://api.niceai.dev/v1', apiKey: process.env.NICEAI_API_KEY })
const resp = await client.chat.completions.create({ model: '<model name>', messages: [{ role: 'user', content: 'Hello' }] })
console.log(resp.choices[0].message.content)
```



## Claude Code

Add this to `~/.claude/settings.json` (Windows: `%USERPROFILE%\.claude\settings.json`):

```json
{
  "env": {
    "ANTHROPIC_BASE_URL": "https://api.niceai.dev",
    "ANTHROPIC_AUTH_TOKEN": "sk-your-key"
  }
}
```

Do not add `/v1`: Claude Code appends `/v1/messages` itself. Details (in Chinese): [Claude Code 接入](/guide/claude-code).

## Codex CLI

Add a custom provider to `~/.codex/config.toml` and set `NICEAI_API_KEY` in your environment:

```toml
model = "<model name>"
model_provider = "niceai"

[model_providers.niceai]
name = "NiceAI"
base_url = "https://api.niceai.dev/v1"
wire_api = "responses"
env_key = "NICEAI_API_KEY"
```

Details (in Chinese): [Codex CLI 接入](/guide/codex).

## Common errors

| Status | Usual cause |
| --- | --- |
| 404 | `/v1` missing or duplicated in the base URL, or a wrong model name |
| 401 | Invalid, incomplete or disabled API key |
| 403 | Insufficient balance, or the key's group cannot use this model |
| 429 | Too many requests: lower concurrency and retry with backoff |

More: [error codes (Chinese)](/guide/error-codes) · [client list (Chinese)](/guide/clients/) · [llms.txt](https://wiki.niceai.dev/llms.txt)
