API Usage Guide

To connect gahorouter to any model SDK, you only need to change two parameters: api_key and base_url.

Requirement: Python ≥ 3.8

Table of Contents

• Get an API Key and bind a group
• Install the SDK
• Model list and rate multipliers
• API usage examples
• Request and response formats
• Supported parameters
• Common errors

Get an API Key and Bind a Group

Log in to https://console.gahorouter.com → "API Keys" → create a key, and you'll get cr-xxx...

In the "Group" column of the API key, click to select a group that matches the model you want to call:

Calling this provider	Must bind this type of group
Claude	Name contains "Claude", "Flagship", or "Premium"
GPT	Name contains "GPT"
Gemini	Name contains "Gemini"
Domestic models	Name contains "DeepSeek / Qwen / GLM / Kimi / MiniMax", etc.

Binding the wrong group will return 401/503; just go back to the console and reselect the group.

Install the SDK

Install only what you need, no need to install everything:

pip install openai anthropic google-generativeai

• openai — call GPT and domestic models
• anthropic — call Claude
• google-generativeai — call Gemini

Model List and Rate Multipliers

The higher the multiplier, the more quota a single call consumes, and the stronger the corresponding capability.

To switch models, just modify the model field in each example.

Claude Series

Group	Model ID
Smart Preferred Channel	claude-haiku-4-5
Official Premium Channel	claude-sonnet-4-6
Official Flagship Channel	claude-opus-4-6

GPT Series

Group	Model ID
GPT-Trial Channel	gpt-5.4

Gemini Series

Group	Model ID
Gemini-Official Channel	gemini-2.5-flash

Domestic Model Series

Group	Model ID
Qwen Trial Channel	qwen-plus
GLM Official Channel (OpenAI interface)	glm-5.1
Kimi Official Channel	kimi-k2.5
MiniMax Trial Channel (OpenAI interface)	MiniMax-M2.5
DeepSeek Trial Channel (OpenAI interface)	deepseek-v3.2

API Usage Examples

Claude

Prerequisite: the Key is bound to a Claude group.

Python:

import anthropic

client = anthropic.Anthropic(
    api_key="cr-xxxxxxxx",
    base_url="https://console.gahorouter.com",
)

msg = client.messages.create(
    model="claude-sonnet-4-6",  # switch model: claude-opus-4-6 / claude-haiku-4-5-20251001
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello"}],
)
print(msg.content[0].text)

curl:

curl https://console.gahorouter.com/v1/messages \
  -H "x-api-key: cr-xxx" \
  -H "anthropic-version: 2023-06-01" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "claude-sonnet-4-6",
  "max_tokens": 1024,
  "messages": [{"role": "user", "content": "Hello"}]
}'

GPT

Prerequisite: the Key is bound to a GPT group. ⚠️ When using the responses interface, you must add stream=True.

Python:

from openai import OpenAI

client = OpenAI(
    api_key="cr-xxxxxxxx",
    base_url="https://console.gahorouter.com/v1",
)

stream = client.responses.create(
    model="gpt-5.4",  # switch model: gpt-4o / gpt-4o-mini / o3 / gpt-5.1-codex-max
    input="Hello",
    stream=True,
)
for ev in stream:
    if ev.type == "response.output_text.delta":
        print(ev.delta, end="", flush=True)

curl:

curl -N https://console.gahorouter.com/v1/responses \
  -H "Authorization: Bearer cr-xxx" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "gpt-5.4",
  "input": "Hello",
  "stream": true
}'

Gemini

Prerequisite: the Key is bound to a Gemini group.

Python:

import google.generativeai as genai

genai.configure(
    api_key="cr-xxxxxxxx",
    client_options={"api_endpoint": "console.gahorouter.com"},
    transport="rest",
)

model = genai.GenerativeModel("gemini-2.5-pro")  # switch model: gemini-2.5-flash
print(model.generate_content("Hello").text)

curl:

curl "https://console.gahorouter.com/v1beta/models/gemini-2.5-pro:generateContent" \
  -H "x-goog-api-key: cr-xxx" \
  -H "Content-Type: application/json" \
  -d '{
  "contents": [{"parts": [{"text": "Hello"}]}]
}'

Domestic Models

Prerequisite: the Key is bound to the corresponding domestic model group.

Python:

from openai import OpenAI

client = OpenAI(
    api_key="cr-xxxxxxxx",
    base_url="https://console.gahorouter.com/v1",
)

resp = client.chat.completions.create(
    model="deepseek-v3.2",  # switch model: qwen-plus / glm-5.1 / kimi-k2.5 / MiniMax-M2.5
    messages=[{"role": "user", "content": "Hello"}],
)
print(resp.choices[0].message.content)

curl:

curl https://console.gahorouter.com/v1/chat/completions \
  -H "Authorization: Bearer cr-xxx" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "deepseek-v3.2",
  "messages": [{"role": "user", "content": "Hello"}]
}'

Request and Response Formats

Claude Format

Request body:

{
  "model": "claude-sonnet-4-6",
  "max_tokens": 1024,
  "system": "You are a helpful assistant.",
  "messages": [
    {"role": "user", "content": "Hello"}
  ]
}

Response body:

{
  "id": "msg_01XFDUDYJgAACzvnptvVoYEL",
  "type": "message",
  "role": "assistant",
  "content": [
    {"type": "text", "text": "Hello! How can I help you?"}
  ],
  "model": "claude-sonnet-4-6",
  "stop_reason": "end_turn",
  "usage": {"input_tokens": 10, "output_tokens": 25}
}

GPT / Domestic Model Format

Request body (chat.completions):

{
  "model": "deepseek-v3.2",
  "messages": [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "Hello"}
  ],
  "temperature": 0.7,
  "max_tokens": 1024,
  "stream": false
}

Response body:

{
  "id": "chatcmpl-abc123",
  "object": "chat.completion",
  "created": 1700000000,
  "model": "deepseek-v3.2",
  "choices": [
    {
      "index": 0,
      "message": {"role": "assistant", "content": "Hello! How can I help you?"},
      "finish_reason": "stop"
    }
  ],
  "usage": {"prompt_tokens": 20, "completion_tokens": 15, "total_tokens": 35}
}

Gemini Format

Request body:

{
  "contents": [
    {"parts": [{"text": "Hello"}]}
  ],
  "generationConfig": {
    "maxOutputTokens": 1024,
    "temperature": 0.7
  }
}

Response body:

{
  "candidates": [
    {
      "content": {"parts": [{"text": "Hello! How can I help you?"}], "role": "model"},
      "finishReason": "STOP"
    }
  ],
  "usageMetadata": {"promptTokenCount": 3, "candidatesTokenCount": 15}
}

Supported Parameters

Claude Interface

Parameter	Type	Required	Description
model	string	Yes	Model ID, see the list above
max_tokens	integer	Yes	Max output tokens, recommended 512–8192
messages	array	Yes	Conversation history, [{"role":"user","content":"..."}]
system	string	No	System prompt
temperature	float	No	Randomness, 0–1, default 1
top_p	float	No	Nucleus sampling, 0–1
top_k	integer	No	Top-K sampling
stream	bool	No	Whether to stream the response, default false
stop_sequences	array	No	Stop generation when these strings are encountered

GPT / Domestic Model Interface

chat.completions

Parameter	Type	Required	Description
model	string	Yes	Model ID
messages	array	Yes	Conversation history, including role (system/user/assistant) and content
temperature	float	No	Randomness, 0–2, default 1
max_tokens	integer	No	Max output tokens
top_p	float	No	Nucleus sampling
stream	bool	No	Whether to stream the response, default false
stop	string/array	No	Stop words
presence_penalty	float	No	Topic diversity, -2–2
frequency_penalty	float	No	Repetition penalty, -2–2
n	integer	No	Number of candidates to generate, default 1

responses (GPT only)

Parameter	Type	Required	Description
model	string	Yes	Model ID
input	string	Yes	User input text
stream	bool	Must be true	Without it, you'll get empty content but still be charged
instructions	string	No	System prompt (equivalent to system)
max_output_tokens	integer	No	Max output tokens
temperature	float	No	Randomness

Gemini Interface

Parameter	Type	Required	Description
contents	array	Yes	Conversation content, [{"parts":[{"text":"..."}]}]
generationConfig.maxOutputTokens	integer	No	Max output tokens
generationConfig.temperature	float	No	Randomness, 0–2
generationConfig.topP	float	No	Nucleus sampling
generationConfig.topK	integer	No	Top-K sampling
systemInstruction	object	No	System prompt, {"parts":[{"text":"..."}]}

Common Errors

Error	Cause	What to Do
401 Unauthorized	Wrong Key	Go back to the console and confirm the Key was copied completely and hasn't expired
403 no group bound	Key not bound to a group	Select a group for the Key in the console
503 / 404	Group does not match the SDK	Claude SDK needs a Claude group; OpenAI responses needs a GPT group; chat.completions needs a domestic model or compatible group
429 Too Many Requests	Upstream momentarily at full capacity	Wait a few seconds and retry; if it persists, ask an administrator to scale up
Empty content returned but still charged	GPT missing stream=True	Add it and it'll work