> ## 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.

# Gemma 4 31B

> A strong low-cost multimodal model on Kyma for image-aware tasks and general work.

## Overview

`gemma-4-31b` is the best low-cost multimodal option in Kyma's active lineup. It is a practical choice when you need image input without moving to a more expensive agent model.

## Specs

| Field              | Value                                        |
| ------------------ | -------------------------------------------- |
| Model ID           | `gemma-4-31b`                                |
| Best for           | Vision, multimodal tasks, cheap general work |
| Context window     | 128K                                         |
| Max output tokens  | 8K                                           |
| Input modalities   | Text, image                                  |
| Output modalities  | Text                                         |
| Tool calling       | Yes                                          |
| Structured outputs | Yes                                          |
| Prompt caching     | Yes                                          |
| Speed              | Medium                                       |
| Cost band          | Cheap                                        |
| Release stage      | Stable                                       |

## Use this when

* You need image input at a lower cost.
* You want a cheap model for extraction or analysis from screenshots/documents.
* You need a general model with decent context and structured outputs.

## Pick something else when

* You want stronger flagship reasoning: use [`qwen-3.6-plus`](/models/qwen-3.6-plus).
* You need stronger multimodal agent behavior: use [`kimi-k2.6`](/models/kimi-k2.6).
* You need 1M context: use `gemini-2.5-flash`.

## 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="gemma-4-31b",
    messages=[{"role": "user", "content": "Describe the key UI problems in this screenshot."}]
)
```