Quick Start

Connect to Nova Router in five minutes, get your API key, call your first model, and connect your tools

Connect to Nova Router in five minutes: get your API key, call your first AI model, and wire it into your development tools or client applications.

Nova Router is a unified AI access gateway compatible with the OpenAI API format. With just one API key and the base URL expected by your client, you can access multiple leading models through the same interface.

1. Three Steps to Get Started in Five Minutes

Step 1 - Get Your API Key

Open https://router.bit-tech.com.cn and enter the console.

1.2 Create a Token

Go to Token Management and create a new API token.

Token List

Create Token

1.3 Copy the URLs and API Key

OpenAI-compatible Base URL: https://router.bit-tech.com.cn/v1
Claude Code Base URL: https://router.bit-tech.com.cn
API Key: sk-...

Important

Save the API key immediately after creation. If you lose it, create a new token.

About Base URLs

Different clients define the base URL differently:

OpenAI-compatible clients such as Codex CLI and Cherry Studio usually use the API endpoint with /v1
Claude Code usually uses the site root URL without /v1

Step 2 - Call Your First API

2.1 Quick Check with curl


curl https://router.bit-tech.com.cn/v1/chat/completions -H "Content-Type: application/json" -H "Authorization: Bearer sk-your-api-key" -d '{
  "model": "gpt-4o",
  "messages": [
    { "role": "user", "content": "Hello, introduce yourself" }
  ]
}'

curl https://router.bit-tech.com.cn/v1/chat/completions -H "Content-Type: application/json" -H "Authorization: Bearer sk-your-api-key" -d '{  "model": "gpt-4o",  "messages": [    { "role": "user", "content": "Hello, introduce yourself" }  ]}'

2.2 Use Python with the OpenAI SDK


from openai import OpenAI

client = OpenAI(
  base_url="https://router.bit-tech.com.cn/v1",
  api_key="sk-your-api-key",
)

response = client.chat.completions.create(
  model="gpt-4o",
  messages=[{"role": "user", "content": "Hello, introduce yourself"}],
)

print(response.choices[0].message.content)

from openai import OpenAIclient = OpenAI(  base_url="https://router.bit-tech.com.cn/v1",  api_key="sk-your-api-key",)response = client.chat.completions.create(  model="gpt-4o",  messages=[{"role": "user", "content": "Hello, introduce yourself"}],)print(response.choices[0].message.content)

2.3 List Available Models


curl https://router.bit-tech.com.cn/v1/models -H "Authorization: Bearer sk-your-api-key"

curl https://router.bit-tech.com.cn/v1/models -H "Authorization: Bearer sk-your-api-key"

Step 3 - Connect Your Tools

Claude Code


export ANTHROPIC_BASE_URL="https://router.bit-tech.com.cn"
export ANTHROPIC_API_KEY="sk-your-api-key"

export ANTHROPIC_BASE_URL="https://router.bit-tech.com.cn"export ANTHROPIC_API_KEY="sk-your-api-key"

Codex CLI


export OPENAI_BASE_URL="https://router.bit-tech.com.cn/v1"
export OPENAI_API_KEY="sk-your-api-key"

export OPENAI_BASE_URL="https://router.bit-tech.com.cn/v1"export OPENAI_API_KEY="sk-your-api-key"

Cherry Studio

Add a provider
Choose the OpenAI-compatible provider type
Fill in the Base URL and API Key
Add the model IDs you want to use

Other OpenAI-Compatible Tools

API URL: https://router.bit-tech.com.cn/v1
API Key: your token
Model name: query it from /v1/models

2. API Capabilities at a Glance

API Endpoint	Description
`POST /v1/chat/completions`	Chat completions
`GET /v1/models`	List available models
`POST /v1/images/generations`	Image generation

For full details, see API Reference.

3. FAQ

How do I see which models are available to me?

Call /v1/models. Each returned id is a model name you can call directly.

How do I troubleshoot `401 Unauthorized`?

First check whether Authorization: Bearer sk-... is set correctly, then confirm that the token is still valid.

How do I troubleshoot `403 Forbidden`?

This is usually caused by model permissions, group permissions, or instance policy restrictions. Confirm that the current token can access the target model.

Where can I find the full parameter reference?

See the complete endpoint documentation in API Reference.

On this page