GitHub Models API - 免费 AI 模型 API 接口服务

pip install openai

from openai import OpenAI

# 初始化客户端
client = OpenAI(
    base_url="https://models.github.ai/inference",
    api_key="YOUR_GITHUB_PAT"  # 替换为您的 GitHub Personal Access Token
)

# 发送请求
response = client.chat.completions.create(
    model="gpt-4o",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "你好，请介绍一下 GitHub Models。"}
    ],
    max_tokens=1000,
    temperature=0.7
)

# 打印响应
print(response.choices[0].message.content)

# 查看 Token 使用情况
print(f"\n使用 Tokens: {response.usage.total_tokens}")
print(f"输入 Tokens: {response.usage.prompt_tokens}")
print(f"输出 Tokens: {response.usage.completion_tokens}")

from openai import OpenAI

client = OpenAI(
    base_url="https://models.github.ai/inference",
    api_key="YOUR_GITHUB_PAT"
)

# 流式输出（适合实时显示）
stream = client.chat.completions.create(
    model="gpt-4o",
    messages=[
        {"role": "user", "content": "写一首关于编程的诗"}
    ],
    stream=True
)

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

print()  # 换行

curl -X POST "https://models.github.ai/inference/chat/completions" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_GITHUB_PAT" \
  -d '{
    "model": "gpt-4o",
    "messages": [
      {
        "role": "system",
        "content": "You are a helpful assistant."
      },
      {
        "role": "user",
        "content": "你好，请介绍一下 GitHub Models。"
      }
    ],
    "max_tokens": 1000,
    "temperature": 0.7
  }'

curl -X POST "https://models.github.ai/inference/chat/completions" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_GITHUB_PAT" \
  -d '{
    "model": "gpt-4o",
    "messages": [
      {"role": "user", "content": "你好"}
    ],
    "stream": true
  }'

npm install openai

import OpenAI from 'openai';

const client = new OpenAI({
  baseURL: 'https://models.github.ai/inference',
  apiKey: process.env.GITHUB_PAT,  // 使用环境变量
});

async function main() {
  const completion = await client.chat.completions.create({
    model: 'gpt-4o',
    messages: [
      { role: 'system', content: 'You are a helpful assistant.' },
      { role: 'user', content: '你好，请介绍一下 GitHub Models。' }
    ],
    max_tokens: 1000,
    temperature: 0.7,
  });

  console.log(completion.choices[0].message.content);
  console.log(`\n使用 Tokens: ${completion.usage.total_tokens}`);
}

main();

# ✅ 精简提示词
messages = [
    {"role": "user", "content": "总结这段文字的要点：[文字内容]"}
]

# ❌ 避免冗余
messages = [
    {"role": "system", "content": "你是一个很厉害的助手，擅长总结..."},
    {"role": "user", "content": "请你帮我总结一下这段文字，注意要抓住重点..."}
]

def log_usage(response):
    """记录 Token 使用情况"""
    usage = response.usage
    print(f"输入: {usage.prompt_tokens} tokens")
    print(f"输出: {usage.completion_tokens} tokens")
    print(f"总计: {usage.total_tokens} tokens")
    
    # 可以保存到文件或数据库
    with open('usage_log.txt', 'a') as f:
        f.write(f"{usage.total_tokens}\n")

from openai import OpenAI
import os

client = OpenAI(
    base_url="https://models.github.ai/inference",
    api_key=os.getenv('GITHUB_PAT')
)

def review_code(code):
    """AI 代码审查"""
    response = client.chat.completions.create(
        model="gpt-4o",
        messages=[
            {"role": "system", "content": "你是一个专业的代码审查专家。"},
            {"role": "user", "content": f"请审查以下代码并提供改进建议：\n\n{code}"}
        ]
    )
    return response.choices[0].message.content

# 使用示例
code = """
def calc(a, b):
    return a + b
"""

review = review_code(code)
print(review)

def generate_docstring(function_code):
    """为函数生成文档字符串"""
    response = client.chat.completions.create(
        model="gpt-4o",
        messages=[
            {"role": "system", "content": "生成 Python 函数的 docstring，使用 Google 风格。"},
            {"role": "user", "content": f"为以下函数生成 docstring：\n\n{function_code}"}
        ]
    )
    return response.choices[0].message.content

# 使用示例
function_code = """
def process_data(data, threshold=0.5):
    filtered = [x for x in data if x > threshold]
    return sum(filtered) / len(filtered) if filtered else 0
"""

docstring = generate_docstring(function_code)
print(docstring)

def compare_models(prompt, models=["gpt-4o", "llama-3.1-70b", "deepseek-r1"]):
    """对比多个模型的输出"""
    results = {}
    
    for model in models:
        try:
            response = client.chat.completions.create(
                model=model,
                messages=[{"role": "user", "content": prompt}]
            )
            results[model] = {
                "response": response.choices[0].message.content,
                "tokens": response.usage.total_tokens
            }
        except Exception as e:
            results[model] = {"error": str(e)}
    
    return results

# 使用示例
prompt = "解释什么是递归，并给出一个 Python 例子。"
comparison = compare_models(prompt)

for model, result in comparison.items():
    print(f"\n{'='*50}")
    print(f"模型: {model}")
    print(f"{'='*50}")
    if "error" in result:
        print(f"错误: {result['error']}")
    else:
        print(result['response'])
        print(f"\n使用 Tokens: {result['tokens']}")