@agentor/server

AI API compatibility layer that exposes AI SDK providers as OpenAI / Anthropic compatible endpoints.

Features

• OpenAI Compatible - Chat Completions, Responses, Completions, Embeddings, Images, Image Edits, Speech, Transcriptions, Rerank, Files, Batches, Models
• Anthropic Compatible - Messages, Message Batches, Models
• Streaming Support - SSE streaming for both OpenAI and Anthropic formats
• Tool Calling - Function calling support for chat completions, messages, and responses
• Multi-Provider - Works with any AI SDK provider via createProviderRegistry
• Batch Processing - OpenAI Batches and Anthropic Message Batches with async processing
• File Management - Upload, list, retrieve, and delete files via OpenAI Files API
• Handler Pattern - Pluggable handler architecture (like unstorage drivers)
• TypeScript-First - Full type safety with official SDK types from openai and @anthropic-ai/sdk

Installation

# Install with npm
$ npm install @agentor/server

# Install with yarn
$ yarn add @agentor/server

# Install with pnpm
$ pnpm add @agentor/server

Quick Start

Setup

import { createServer, openaiHandler, anthropicHandler } from "@agentor/server";
import { createProviderRegistry } from "ai";
import { createDashScope } from "@agentor/dashscope";

const dashscope = createDashScope({ apiKey: process.env.DASHSCOPE_API_KEY });

const server = createServer({
  registry: createProviderRegistry({ dashscope }),
  handlers: [openaiHandler(), anthropicHandler()],
  models: [
    {
      id: "dashscope:qwen3.5-flash",
      display_name: "Qwen 3.5 Flash",
      created: 1747526400,
      owned_by: "aliyun",
    },
  ],
});

server.listen(3000);

OpenAI Client

import OpenAI from "openai";

const client = new OpenAI({ baseURL: "http://localhost:3000/v1", apiKey: "any" });

// Chat completion
const response = await client.chat.completions.create({
  model: "dashscope:qwen3.5-flash",
  messages: [{ role: "user", content: "Hello" }],
});

console.log(response.choices[0].message.content);

// Streaming
const stream = await client.chat.completions.create({
  model: "dashscope:qwen3.5-flash",
  messages: [{ role: "user", content: "Hello" }],
  stream: true,
});

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

Anthropic Client

import Anthropic from "@anthropic-ai/sdk";

const client = new Anthropic({ baseURL: "http://localhost:3000", apiKey: "any" });

// Message
const message = await client.messages.create({
  model: "dashscope:qwen3.5-flash",
  max_tokens: 1024,
  messages: [{ role: "user", content: "Hello" }],
});

console.log(message.content[0].text);

// Streaming
const stream = client.messages.stream({
  model: "dashscope:qwen3.5-flash",
  max_tokens: 1024,
  messages: [{ role: "user", content: "Hello" }],
});

for await (const event of stream) {
  if (event.type === "content_block_delta" && event.delta.type === "text_delta") {
    process.stdout.write(event.delta.text);
  }
}

Endpoints

OpenAI Handler

| Endpoint | Method | Description | | -------------------------- | ------ | ------------------------------------- | | /v1/chat/completions | POST | Chat completion (stream + non-stream) | | /v1/responses | POST | Responses API (stream + non-stream) | | /v1/completions | POST | Legacy completions | | /v1/embeddings | POST | Text embeddings | | /v1/images/generations | POST | Image generation | | /v1/images/edits | POST | Image editing | | /v1/audio/speech | POST | Text-to-speech | | /v1/audio/transcriptions | POST | Speech-to-text | | /v1/audio/translations | POST | Audio translation | | /v1/rerank | POST | Document reranking | | /v1/files | POST | Upload file | | /v1/files | GET | List files | | /v1/files/:id | GET | Retrieve file metadata | | /v1/files/:id/content | GET | Download file content | | /v1/files/:id | DELETE | Delete file | | /v1/batches | POST | Create batch | | /v1/batches | GET | List batches | | /v1/batches/:id | GET | Retrieve batch status | | /v1/batches/:id/cancel | POST | Cancel batch | | /v1/models | GET | List available models |

Anthropic Handler

| Endpoint | Method | Description | | ---------------------------------- | ------ | -------------------------------------- | | /v1/messages | POST | Message creation (stream + non-stream) | | /v1/messages/batches | POST | Create message batch | | /v1/messages/batches | GET | List message batches | | /v1/messages/batches/:id | GET | Retrieve message batch status | | /v1/messages/batches/:id/cancel | POST | Cancel message batch | | /v1/messages/batches/:id | DELETE | Delete message batch | | /v1/messages/batches/:id/results | GET | Get message batch results (JSONL) | | /v1/models | GET | List available models |

Models Configuration

The models option accepts string shortcuts or full objects:

createServer({
  // ...
  models: [
    // String shortcut — minimal defaults
    "dashscope:qwen3.5-flash",
    // Full object — control all fields
    {
      id: "dashscope:qwen-plus",
      display_name: "Qwen Plus",
      created: 1704067200,
      owned_by: "aliyun",
    },
  ],
});

Handler Options

Both handlers are mounted at /v1 by default:

import { openaiHandler, anthropicHandler } from "@agentor/server";

createServer({
  registry: createProviderRegistry({ dashscope }),
  handlers: [openaiHandler(), anthropicHandler()],
});

Programmatic Usage

Use app.fetch() for testing without starting a real HTTP server:

import { createServer, openaiHandler } from "@agentor/server";

const { app } = createServer({
  registry: createProviderRegistry({ dashscope }),
  handlers: [openaiHandler()],
});

const response = await app.fetch(
  new Request("http://localhost/v1/chat/completions", {
    method: "POST",
    headers: { "Content-Type": "application/json" },
    body: JSON.stringify({
      model: "dashscope:qwen3.5-flash",
      messages: [{ role: "user", content: "Hello" }],
    }),
  }),
);

console.log(await response.json());

Supported Parameters

OpenAI → AI SDK

| OpenAI Parameter | AI SDK Parameter | | ---------------- | ---------------- | | temperature | temperature | | top_p | topP | | max_tokens | maxTokens | | stop | stopSequences | | stream | streamText | | tools | tools |

Anthropic → AI SDK

| Anthropic Parameter | AI SDK Parameter | | ------------------- | ---------------- | | system | system | | max_tokens | maxTokens | | temperature | temperature | | top_p | topP | | stream | streamText | | tools | tools | | stop_sequences | stopSequences |

License

MIT © Demo Macro