DeepSeek
DeepSeek V4 Flash
deepseek/deepseek-v4-flash
Very cheap text model with a 1M context for fast inference.
TextText → TextReady on endue
- Low cost
- Fast replies
- Long documents
Overview
Good for
Written by endue
- Low cost
- Fast replies
- Long documents
Not ideal for
From the model’s specs
- Text input only. Images and files you attach are skipped.
Capabilities
| Tool use | Supported |
|---|---|
| Structured output | Supported |
| JSON mode | Supported |
| Reasoning | Supported |
| Vision | Not supported |
| Audio input | Not supported |
| Video input | Not supported |
| File input | Not supported |
| Long context | Supported |
Specifications
- Model ID
deepseek/deepseek-v4-flash- Family
- DeepSeek V4
- Context window
- 1M tokens
- Input
- Text
- Output
- Text
- Released
- 2026-04-24