DeepSeek: DeepSeek V4 Flash Vision Exp

Section ID: DOC-MODEL-k0td6kk2qo#overview

deepseek — deepseek/deepseek-v4-flash-vision-exp

Input modalities: Reported — text, image

Output modalities: Reported — text

Context: 1048576 tokens

Input limit: Unknown

Output limit: Unknown

Input character limit: Unknown

Tool calling: Supported

Structured output: Supported

Reasoning: Supported

Streaming: Unknown

Prompt caching: Unknown

New selection: Undecided

Existing saved use: Undecided

All available models

Document ID: DOC-MODEL-k0td6kk2qo