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

# DeepSeek V3

> Kyma's best value frontier-class model for general work, coding, and analysis.

## Overview

`deepseek-v3` is the value pick in Kyma's catalog. It gives you strong reasoning and coding quality without pushing you into the highest cost tier.

## Specs

| Field              | Value                                  |
| ------------------ | -------------------------------------- |
| Model ID           | `deepseek-v3`                          |
| Best for           | General work, coding, reasoning, value |
| Context window     | 160K                                   |
| Max output tokens  | 8K                                     |
| Input modalities   | Text                                   |
| Output modalities  | Text                                   |
| Tool calling       | Yes                                    |
| Structured outputs | Yes                                    |
| Prompt caching     | No                                     |
| Speed              | Medium                                 |
| Cost band          | Balanced                               |
| Release stage      | Stable                                 |

## Use this when

* You want a cheaper alternative to flagship models.
* You need one model that still handles both coding and analysis well.
* You care about value more than brand or benchmark prestige.

## Pick something else when

* You want the strongest default quality: use [`qwen-3.6-plus`](/models/qwen-3.6-plus).
* You need deeper reasoning specifically: use [`deepseek-r1`](/models/deepseek-r1).
* You need multimodal or vision input: use [`kimi-k2.6`](/models/kimi-k2.6) or `gemma-4-31b`.

## Example

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

client = OpenAI(base_url="https://kymaapi.com/v1", api_key="ky-...")

response = client.chat.completions.create(
    model="deepseek-v3",
    messages=[{"role": "user", "content": "Propose three architectures for this internal copilot and compare tradeoffs."}]
)
```