# DeepSeek V4 Flash Vision Exp

Experimental vision-enabled version of DeepSeek V4 Flash 0731. Sparse mixture-of-experts model (13B active of 284B total) accepting text and image inputs for visual question answering, document/chart understanding, and multimodal agent workflows.

## Overview

- **Provider:** DeepSeek
- **Model ID:** `deepseek-v4-flash-vision-exp`
- **Type:** Chat
- **Canonical URL:** https://modelstack.cc/models/chat/deepseek-v4-flash-vision-exp
- **Version:** V4 Flash Vision Exp
- **Best for:** Document and chart understanding, visual question answering, multimodal agent workflows, cost-effective vision reasoning

## Pricing

| Usage | Price |
| --- | ---: |
| Input | $0.22 / 1M tokens |
| Output | $0.66 / 1M tokens |
| Cache read | $0.007 / 1M tokens |
| Cache write (5 minutes) | $0.22 / 1M tokens |

## Specifications

- **Context window:** 1M
- **Maximum output:** 384K
- **Released:** 2026-08-21

## Capabilities

- chat
- tool_use
- reasoning
- coding
- analysis

## API usage

Use this model ID in a ModelStack API request:

```json
{
  "model": "deepseek-v4-flash-vision-exp"
}
```

For the rendered page, visit https://modelstack.cc/models/chat/deepseek-v4-flash-vision-exp.