Llama 4 Maverick API 完整指南：Meta 开源大模型 2026 开发者教程

import openai

client = openai.OpenAI(
    api_key="你的-crazyrouter-api-key",
    base_url="https://crazyrouter.com/v1"
)

response = client.chat.completions.create(
    model="meta-llama/llama-4-maverick",
    messages=[
        {"role": "system", "content": "你是一个专业的AI助手，擅长中文回答。"},
        {"role": "user", "content": "解释一下 Transformer 架构中的自注意力机制。"}
    ],
    max_tokens=2000,
    temperature=0.7
)

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

import OpenAI from 'openai';

const client = new OpenAI({
  apiKey: '你的-crazyrouter-api-key',
  baseURL: 'https://crazyrouter.com/v1'
});

const response = await client.chat.completions.create({
  model: 'meta-llama/llama-4-maverick',
  messages: [
    { role: 'system', content: '你是一个专业的编程助手。' },
    { role: 'user', content: '用 Python 实现一个简单的 LRU 缓存。' }
  ],
  max_tokens: 2000,
  stream: true
});

for await (const chunk of response) {
  process.stdout.write(chunk.choices[0]?.delta?.content || '');
}

curl https://crazyrouter.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer 你的-crazyrouter-api-key" \
  -d '{
    "model": "meta-llama/llama-4-maverick",
    "messages": [
      {"role": "user", "content": "什么是 Llama 4 Maverick？"}
    ],
    "max_tokens": 1000
  }'

# 安装 vLLM
pip install vllm

# 启动 Llama 4 Maverick 服务（需要 8xA100 80GB）
python -m vllm.entrypoints.openai.api_server \
  --model meta-llama/Llama-4-Maverick-17B-128E \
  --tensor-parallel-size 8 \
  --max-model-len 131072 \
  --port 8000