Tencent Hunyuan API Complete Guide: Python Integration and WeChat Ecosystem

What This Tutorial Covers

You will master the complete Tencent Hunyuan LLM API:

• Tencent Cloud platform registration and API Key acquisition
• Hunyuan Turbo / Hunyuan Lite model selection
• Seamless OpenAI compatibility mode switching
• Function Calling and tool integration
• WeChat Mini Program / WeCom integration
• Production cost optimization

🎯 Tencent Hunyuan is a major player in China’s LLM space. Its biggest advantage: seamless integration with the WeChat ecosystem (Mini Programs / Official Accounts / WeCom).

---

Step 1: Meet the Hunyuan Model Family

Model	Context	Max Output	Characteristics	Price (per M tokens)
Hunyuan Turbo	32K	8K	Flagship, high-performance reasoning	¥15 / ¥50
Hunyuan Pro	32K	8K	Balanced performance	¥10 / ¥30
Hunyuan Standard	8K	4K	Daily tasks	¥3 / ¥9
Hunyuan Lite	8K	4K	Free, best for getting started	Free
Hunyuan Vision	32K	8K	Vision understanding	Pay-as-you-go
Hunyuan Embedding	—	—	Text vectorization	¥0.7 / M tokens

---

Step 2: Get Your API Key

1. Visit the Tencent Cloud Console and sign up
2. Activate the Hunyuan LLM service
3. Create keys under Access Management > API Key Management
4. Obtain your SecretId + SecretKey

export TENCENT_SECRET_ID="AKIDxxxxxx"
export TENCENT_SECRET_KEY="xxxxxxxxxx"

---

Step 3: OpenAI Compatibility Mode (Recommended)

Hunyuan is fully compatible with the OpenAI SDK — one-line switch:

pip install openai

from openai import OpenAI
import os

client = OpenAI(
    api_key=os.getenv("TENCENT_SECRET_ID"),  # Use SecretId as the API Key
    base_url="https://api.hunyuan.cloud.tencent.com/v1",
)

response = client.chat.completions.create(
    model="hunyuan-turbo",
    messages=[
        {"role": "system", "content": "You are a professional cloud architect"},
        {"role": "user", "content": "Design a cloud architecture for an e-commerce app with 1M DAU"}
    ],
    temperature=0.7,
    max_tokens=2048,
)

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

Streaming Output

response = client.chat.completions.create(
    model="hunyuan-turbo",
    messages=[{"role": "user", "content": "Introduce Tencent Cloud's main product lines"}],
    stream=True,
)

for chunk in response:
    if chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="", flush=True)

---

Step 4: Function Calling

import json

tools = [{
    "type": "function",
    "function": {
        "name": "query_cloud_resources",
        "description": "Query resource usage under a Tencent Cloud account",
        "parameters": {
            "type": "object",
            "properties": {
                "resource_type": {
                    "type": "string",
                    "enum": ["cvm", "cdn", "cos", "cdn"],
                    "description": "Resource type"
                },
                "region": {"type": "string", "description": "Region, e.g., ap-guangzhou"}
            },
            "required": ["resource_type"]
        }
    }
}]

def query_cloud_resources(resource_type: str, region: str = "ap-guangzhou") -> dict:
    """Simulated Tencent Cloud resource query (call Tencent Cloud API in production)"""
    resources = {
        "cvm": {"instances": 12, "total_cpu": 48, "total_memory": "192GB"},
        "cdn": {"domains": 8, "total_bandwidth": "20Gbps"},
        "cos": {"buckets": 15, "total_storage": "5.2TB"},
    }
    return resources.get(resource_type, {"error": "Unknown resource"})

def chat_with_cloud_assistant(query: str) -> str:
    messages = [{"role": "user", "content": query}]

    response = client.chat.completions.create(
        model="hunyuan-turbo",
        messages=messages,
        tools=tools,
        tool_choice="auto",
    )

    msg = response.choices[0].message

    if not msg.tool_calls:
        return msg.content

    # Handle tool calls
    messages.append(msg)

    for tool_call in msg.tool_calls:
        func_name = tool_call.function.name
        func_args = json.loads(tool_call.function.arguments)
        result = query_cloud_resources(**func_args)

        messages.append({
            "role": "tool",
            "tool_call_id": tool_call.id,
            "content": json.dumps(result, ensure_ascii=False)
        })

    final = client.chat.completions.create(
        model="hunyuan-turbo",
        messages=messages,
    )

    return final.choices[0].message.content

print(chat_with_cloud_assistant("How are our CVM and CDN resources doing?"))

---

Step 5: WeChat Ecosystem Integration

Hunyuan’s biggest differentiator: deep integration with WeChat.

WeChat Mini Program Integration

// Mini Program client code
Page({
  async askAI() {
    const res = await wx.cloud.callContainer({
      config: { env: 'prod-xxx' },
      path: '/hunyuan/chat',
      method: 'POST',
      data: {
        model: 'hunyuan-lite',
        messages: [{ role: 'user', content: this.data.userInput }],
      },
    })
    this.setData({ answer: res.data.choices[0].message.content })
  }
})

WeCom Group Bot

# WeCom group bot + Hunyuan AI
from flask import Flask, request
import requests

app = Flask(__name__)
WECOM_WEBHOOK = "https://qyapi.weixin.qq.com/cgi-bin/webhook/send?key=xxx"

@app.route("/wecom-bot", methods=["POST"])
def wecom_bot():
    data = request.json
    user_msg = data.get("text", {}).get("content", "")

    # Call Hunyuan AI for a reply
    response = client.chat.completions.create(
        model="hunyuan-lite",
        messages=[{"role": "user", "content": user_msg}],
        max_tokens=512,
    )

    ai_reply = response.choices[0].message.content

    # Send back to WeCom group
    requests.post(WECOM_WEBHOOK, json={
        "msgtype": "text",
        "text": {"content": f"🤖 AI Assistant: {ai_reply}"}
    })

    return "ok"

if __name__ == "__main__":
    app.run(port=8080)

---

Step 6: Tencent Cloud SDK Native Call

pip install tencentcloud-sdk-python

from tencentcloud.common import credential
from tencentcloud.hunyuan.v20230901 import hunyuan_client, models

cred = credential.Credential(
    os.getenv("TENCENT_SECRET_ID"),
    os.getenv("TENCENT_SECRET_KEY"),
)

client = hunyuan_client.HunyuanClient(cred, "ap-guangzhou")

req = models.ChatCompletionsRequest()
req.Model = "hunyuan-turbo"
req.Messages = [
    {"Role": "user", "Content": "Introduce Tencent Hunyuan LLM in three sentences"}
]

resp = client.ChatCompletions(req)
print(resp.Choices[0].Message.Content)

---

Step 7: Cost Optimization

Model Selection Guide

Scenario	Model	Daily Cost Estimate (10K calls)
Dev/Test	Hunyuan Lite	¥0 (free)
Customer service bot	Hunyuan Standard	~¥3
Content generation	Hunyuan Pro	~¥10
Complex analysis	Hunyuan Turbo	~¥15

Optimization Tips

# 1. Use Lite for short tasks
def quick_reply(msg: str) -> str:
    response = client.chat.completions.create(
        model="hunyuan-lite",  # Free model
        messages=[{"role": "user", "content": msg}],
        max_tokens=128,         # Limit output
    )
    return response.choices[0].message.content

# 2. Only use Turbo for complex tasks
def deep_analysis(doc: str) -> str:
    response = client.chat.completions.create(
        model="hunyuan-turbo",
        messages=[{"role": "user", "content": f"In-depth analysis: {doc}"}],
        max_tokens=2048,
    )
    return response.choices[0].message.content

---

Comparison with Other Models

Dimension	Hunyuan Turbo	ERNIE 4.5	Qwen3
WeChat integration	⭐⭐⭐⭐⭐	⭐⭐	⭐⭐
Chinese comprehension	⭐⭐⭐⭐	⭐⭐⭐⭐⭐	⭐⭐⭐⭐
API pricing	Moderate	Moderate	Moderate
Tencent Cloud ecosystem	⭐⭐⭐⭐⭐	—	—
Coding ability	⭐⭐⭐	⭐⭐⭐	⭐⭐⭐⭐

If you’re within the Tencent Cloud ecosystem (Mini Programs / WeCom / CloudBase), Hunyuan is the natural first choice.

---

FAQ

Q: Can Hunyuan integrate with other Tencent AI services?

A: Yes. Hunyuan + Tencent Cloud Document Parsing + Tencent Cloud Translation + Tencent Cloud Speech Recognition can build a complete AI application pipeline.

Q: Is the free model sufficient?

A: Hunyuan Lite is fully adequate for daily customer service, simple Q&A, and copy generation — and it’s permanently free.

Q: Can it integrate with WeChat Official Accounts?

A: Yes. Through the Official Account developer mode + Hunyuan API, you can implement AI auto-reply.

---

Next Steps

• Baidu ERNIE API Guide
• China AI Models Ultimate Comparison

📝 Based on Tencent Hunyuan latest API version tested in June 2026.