AI Hub
← All models

DeepSeek-V4-Flash vs Qwen3 VL 4B

DeepSeek vs Alibaba — benchmarks, pricing, and capabilities side by side.

  • DeepSeek-V4-Flash has the higher intelligence index (89.4 vs 44.3)
  • Qwen3 VL 4B is cheaper ($0.00 vs $0.10 per 1M input)
DeepSeek-V4-FlashQwen3 VL 4B
Intelligence index89.444.3
DeveloperDeepSeekAlibaba
TypeLLMLLM
AccessOpen weights
Context window1,048,576 tokens
Input price$0.10 / 1M$0.00 / 1M
Output price$0.20 / 1M$0.00 / 1M
Speed109 tok/s
ReleasedApril 24, 2026October 14, 2025
Parameters284B (13B active)
Input modalitiesText
Output modalitiesText

Shared benchmarks

DeepSeek-V4-Flash
Qwen3 VL 4B
GPQA Diamond
89.4
49.4
Humanity’s Last Exam
32.1
4.4