CoreRouter API 文档Node.js SDK 接入
使用 OpenAI 官方 Node.js SDK 时，只需要把 apiKey、baseURL 和 model 换成 CoreRouter 信息。
安装
bash
npm install openai
也可以使用 Bun：
bash
bun add openai
如果要直接读取 .env 文件，再安装 dotenv：
bash
npm install dotenv
或：
bash
bun add dotenv
基础使用
typescript
import OpenAI from "openai";
const client = new OpenAI({
apiKey: process.env.COREROUTER_API_KEY,
baseURL: "https://api.corerouter.tech/v1",
});
const completion = await client.chat.completions.create({
model: "claude-sonnet-4-5",
messages: [
{ role: "user", content: "你好，请介绍一下你自己。" },
],
});
console.log(completion.choices[0].message.content);
流式输出
typescript
import OpenAI from "openai";
const client = new OpenAI({
apiKey: process.env.COREROUTER_API_KEY,
baseURL: "https://api.corerouter.tech/v1",
});
const stream = await client.chat.completions.create({
model: "claude-sonnet-4-5",
messages: [
{ role: "user", content: "用三句话介绍 CoreRouter。" },
],
stream: true,
});
for await (const chunk of stream) {
process.stdout.write(chunk.choices[0]?.delta?.content ?? "");
}
多轮对话
typescript
import type { ChatCompletionMessageParam } from "openai/resources/chat/completions";
const messages: ChatCompletionMessageParam[] = [
{ role: "user", content: "你好" },
];
let completion = await client.chat.completions.create({
model: "claude-sonnet-4-5",
messages,
});
messages.push({
role: "assistant",
content: completion.choices[0].message.content ?? "",
});
messages.push({ role: "user", content: "请写一个 TypeScript hello 函数。" });
completion = await client.chat.completions.create({
model: "claude-sonnet-4-5",
messages,
});
console.log(completion.choices[0].message.content);
工具调用
typescript
const completion = await client.chat.completions.create({
model: "claude-sonnet-4-5",
messages: [{ role: "user", content: "北京天气怎么样？" }],
tools: [
{
type: "function",
function: {
name: "get_weather",
description: "获取指定城市的天气信息",
parameters: {
type: "object",
properties: {
city: { type: "string", description: "城市名称" },
},
required: ["city"],
},
},
},
],
});
const toolCall = completion.choices[0].message.tool_calls?.[0];
if (toolCall) {
console.log(toolCall.function.name);
console.log(toolCall.function.arguments);
}
Responses API
这段示例要求 Model ID 支持 Responses。普通聊天模型不一定可用。 如果 client.responses 不存在，请先升级 openai 包，或者直接按 Responses API 的 curl 示例调用。
typescript
const response = await client.responses.create({
model: "responses-model-id",
input: "请用三句话介绍 CoreRouter。",
});
console.log(response.output_text);
Embeddings
typescript
const result = await client.embeddings.create({
model: "embedding-model-id",
input: ["第一段文本", "第二段文本"],
});
console.log(result.data[0].embedding.length);
使用 .env
bash
COREROUTER_API_KEY=sk-xxxxxxxxxxxxxxxx
COREROUTER_BASE_URL=https://api.corerouter.tech/v1
COREROUTER_MODEL=claude-sonnet-4-5
typescript
import "dotenv/config";
import OpenAI from "openai";
const client = new OpenAI({
apiKey: process.env.COREROUTER_API_KEY,
baseURL: process.env.COREROUTER_BASE_URL,
});
const completion = await client.chat.completions.create({
model: process.env.COREROUTER_MODEL ?? "claude-sonnet-4-5",
messages: [{ role: "user", content: "Hello" }],
});
console.log(completion.choices[0].message.content);
请把 .env 加入 .gitignore，不要把 API Key 提交到仓库。
常见问题
COREROUTER_API_KEY为空：确认从同一个终端启动程序，或部署平台已配置环境变量。- 返回
model not found：确认model是控制台中的 Model ID。 - 普通对话可用但工具调用失败：换用支持 Tool Calling 的模型。