Back to Models

DeepSeek: DeepSeek V4 Flash Vision Exp

Deepseektext
9.1 / 10 Overall Rating

DeepSeek V4 Flash Vision Exp is an experimental multimodal model by DeepSeek. It extends the DeepSeek V4 Flash 0731 text model with image understanding capabilities, maintaining its existing text performance, and features an industry-leading 1M token context window at highly competitive pricing.

Context Window
1049K
Knowledge Cutoff
2025
Max Output
16K
License
Proprietary

Where it Excels (Pros)

  • Massive 1M token context window
  • Supports 384K output tokens
  • Significantly cheaper than competitors
  • Advanced multimodal image understanding
  • Full support for agentic workflows

Limitations (Cons)

  • Uncertain long-term model stability
  • Unverified performance against key baselines
  • Unproven for complex long-form writing

Benchmark Breakdown

Reasoning (MMLU)82%
Coding (HumanEval)78%
Mathematics (MATH)76%
Long-Context Retrieval95%

Pricing Matrix

Input cost / 1M tokens$0.22
Output cost / 1M tokens$0.66

Compare DeepSeek: DeepSeek V4 Flash Vision Exp with other frontier models:

Related Skills, Tools & Automations