HAI Gateway Docs

DeepSeek

DeepSeek Chat Completions

Call DeepSeek models via the OpenAI-compatible endpoint, supporting streaming, tool calling, thinking, and JSON mode.

POSTv1/chat/completions

Authorizations

Authorizationstringheaderrequired

OpenAI compatible endpoint: Bearer sk-pat-YourAccessToken

Use your HAI Gateway personal access token (starting with sk-pat-). No separate DeepSeek API key is required.

Base URL

https://api.hai.network/compatible-preview/openai

Body Parameters

The request body follows the OpenAI Chat Completions API format. For detailed parameter descriptions, refer to the DeepSeek official documentation.

DeepSeek Notes
  • Only Chat Completion endpoints are supported: DeepSeek does not support the OpenAI Responses (/v1/responses) endpoint. Please use /v1/chat/completions or Anthropic /v1/messages.
  • Thinking is only available on pro: enable_thinking (OpenAI endpoint) or thinking (Anthropic endpoint) only works with deepseek-v4-pro; deepseek-v4-flash ignores these parameters and does not return thinking content.
  • Thinking field location: On the OpenAI endpoint, the reasoning chain is in the choices[].message.reasoning_content field; on the Anthropic endpoint, it appears as a thinking block in the content array.
  • Do not pass reasoning in multi-turn context: When building conversation history, only pass back the previous turn's content (OpenAI) or text block (Anthropic) — do not include reasoning_content / thinking blocks in messages, otherwise a parameter error will be returned.
  • developer role not supported: The OpenAI compatible endpoint only accepts system / user / assistant roles; the Anthropic endpoint follows the official user / assistant conversation structure, with system messages passed via the top-level system field.
cURL
cURL:Default
curl https://api.hai.network/compatible-preview/openai/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-pat-YOUR_ACCESS_TOKEN" \
  -d '{
    "model": "deepseek-v4-flash",
    "messages": [
      {
        "role": "user",
        "content": "Hello, introduce yourself in one sentence"
      }
    ]
  }'
200
200:Default
{
  "id": "chatcmpl-...",
  "object": "chat.completion",
  "model": "deepseek-v4-flash",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "I am DeepSeek, a large language model developed by DeepSeek."
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 14,
    "completion_tokens": 20,
    "total_tokens": 34
  }
}