DWA

A single Go binary that exposes the OpenAI and Anthropic chat APIs on top of the DeepSeek web app.

Real capture: the live server answering curl calls to both endpoints.

Overview

DWA talks to the unofficial chat.deepseek.com web API and re-exposes it through two standard interfaces:

Point any OpenAI or Anthropic SDK at http://127.0.0.1:8000 and it works. Streaming and non-streaming are both supported, and the proof-of-work challenge the web app requires is solved locally with the original WASM module.

How it works

Your clientOpenAI / Anthropic SDK
DWAtranslate + solve POW
chat.deepseek.comSSE stream

Each request fetches a POW challenge and a HIF token in parallel, solves the challenge through sha3_pow.wasm, then streams the completion back and converts it to the format the client asked for. Conversations reuse a cached DeepSeek session so multi-turn chats keep their context.

Requirements

  Setup: 1. Get your token

The token is the Bearer value the web app sends with each request. It is not a cookie, so you read it from the request headers.

1

Log in at chat.deepseek.com and send any message.

2

Open DevTools with F12, go to the Network tab, and click the request named completion.

3

Under Headers, copy the value after Authorization: Bearer.

DevTools / Network / Headers
DevTools Headers panel showing the Authorization Bearer token outlined in red
The value boxed in red is your token. Copy everything after Bearer.
The token belongs to your session and expires over time. When requests start failing with 40003 invalid token, grab a fresh one the same way.

  Setup: 2. Configure

Copy the template and paste your token in:

cp .env.example .env

Then edit .env:

DEEPSEEK_TOKEN=your_token_here
HOST=127.0.0.1
PORT=8000
.env is gitignored, so your token never lands in version control.

  Setup: 3. Run

From the project root:

go run .

On Windows you can double-click start.bat instead. It builds the binary into bin/ and launches it.

bash
$ go run .
 
DWA - DeepSeek Web API
OpenAI: http://127.0.0.1:8000/v1/chat/completions
Anthropic: http://127.0.0.1:8000/v1/messages
Models: http://127.0.0.1:8000/v1/models
Server output once it is up.

  Endpoints

MethodPathFormatNotes
GET/v1/models-Lists available model IDs
POST/v1/chat/completionsOpenAIStreaming and non-streaming
POST/v1/messagesAnthropicStreaming and non-streaming

Examples

OpenAI, non-streaming

curl http://127.0.0.1:8000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{"model":"deepseek-chat","messages":[{"role":"user","content":"Say hi in one word"}]}'
{
  "id": "chatcmpl-18b4adb1e8094f38",
  "object": "chat.completion",
  "model": "deepseek-chat",
  "choices": [
    {
      "index": 0,
      "message": { "role": "assistant", "content": "Hi" },
      "finish_reason": "stop"
    }
  ],
  "usage": { "prompt_tokens": 0, "completion_tokens": 0, "total_tokens": 0 }
}

OpenAI, streaming

Add "stream": true. Chunks arrive as server-sent events:

data: {"choices":[{"delta":{"content":"1"},"index":0}],"object":"chat.completion.chunk"}

data: {"choices":[{"delta":{"content":" "},"index":0}],"object":"chat.completion.chunk"}

data: {"choices":[{"delta":{"content":"2"},"index":0}],"object":"chat.completion.chunk"}

data: [DONE]

Anthropic

curl http://127.0.0.1:8000/v1/messages \
  -H "Content-Type: application/json" \
  -d '{"model":"deepseek-chat","max_tokens":100,"messages":[{"role":"user","content":"Say hello in one word"}]}'
{
  "id": "msg_18b4adb4b6943064",
  "type": "message",
  "role": "assistant",
  "model": "deepseek-chat",
  "content": [{ "type": "text", "text": "Hello" }],
  "stop_reason": "end_turn",
  "usage": { "input_tokens": 1, "output_tokens": 1 }
}

Configuration

VariableDefaultDescription
DEEPSEEK_TOKEN(required)Bearer token from the web app
HOST127.0.0.1Bind address
PORT8000Bind port

Flags override the environment:

go run . -host 0.0.0.0 -port 9000

Notes and limitations