> ## Documentation Index
> Fetch the complete documentation index at: https://api-docs.upmore.net/llms.txt
> Use this file to discover all available pages before exploring further.

# DeepSeek V4

> DeepSeek V4 reasoning chat models (flash / pro), OpenAI-compatible.

The DeepSeek V4 models are available through Upmore via the standard Chat Completions API (`/v1/chat/completions`). Both are reasoning models — they think before answering and stream a `reasoning_content` field alongside the reply.

## Available models

| Model               | Notes                                                                         |
| ------------------- | ----------------------------------------------------------------------------- |
| `deepseek-v4-pro`   | Flagship — strongest reasoning and coding                                     |
| `deepseek-v4-flash` | Light — low latency and cost, good for everyday chat and high-concurrency use |

<Note>
  Reasoning tokens are billed as **completion tokens**. Set `max_tokens` generously — the thinking phase consumes part of the budget before the visible answer starts.
</Note>

## Quick example

<CodeGroup>
  ```bash cURL theme={null}
  curl https://api.upmore.net/v1/chat/completions \
    -H "Authorization: Bearer YOUR_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
      "model": "deepseek-v4-pro",
      "messages": [
        { "role": "user", "content": "Explain quantum entanglement in one paragraph." }
      ]
    }'
  ```

  ```python Python theme={null}
  from openai import OpenAI

  client = OpenAI(
      api_key="YOUR_API_KEY",
      base_url="https://api.upmore.net/v1"
  )

  response = client.chat.completions.create(
      model="deepseek-v4-pro",
      messages=[
          {"role": "user", "content": "Explain quantum entanglement in one paragraph."}
      ]
  )

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

  ```python Streaming theme={null}
  from openai import OpenAI

  client = OpenAI(
      api_key="YOUR_API_KEY",
      base_url="https://api.upmore.net/v1"
  )

  stream = client.chat.completions.create(
      model="deepseek-v4-flash",
      messages=[
          {"role": "user", "content": "Write a short poem about the sea."}
      ],
      stream=True
  )

  for chunk in stream:
      delta = chunk.choices[0].delta if chunk.choices else None
      if delta and getattr(delta, "reasoning_content", None):
          print(delta.reasoning_content, end="")   # thinking phase
      elif delta and delta.content:
          print(delta.content, end="")             # final answer
  ```
</CodeGroup>

## Parameters

| Parameter     | Type           | Required | Description                                      |
| ------------- | -------------- | -------- | ------------------------------------------------ |
| `model`       | string         | Yes      | `deepseek-v4-pro` or `deepseek-v4-flash`         |
| `messages`    | array          | Yes      | List of `{ role, content }` objects              |
| `stream`      | boolean        | No       | Enable SSE streaming. Default: `false`           |
| `temperature` | float          | No       | Controls randomness. Default: `1`                |
| `top_p`       | float          | No       | Nucleus sampling threshold. Default: `1`         |
| `max_tokens`  | integer        | No       | Max output tokens **including reasoning tokens** |
| `stop`        | string / array | No       | Sequences that stop generation                   |
| `tools`       | array          | No       | List of tools the model may call                 |

## Response notes

| Field                                              | Description                                                          |
| -------------------------------------------------- | -------------------------------------------------------------------- |
| `choices[].message.reasoning_content`              | The model's chain-of-thought (streamed as `delta.reasoning_content`) |
| `usage.completion_tokens_details.reasoning_tokens` | Reasoning tokens spent — part of `completion_tokens`                 |
| `usage.prompt_tokens_details.cached_tokens`        | Prompt cache hits                                                    |

<Card title="API Reference" icon="code" href="/api-reference/model-api/deepseek/v4">
  View the interactive API playground.
</Card>