[KongchangAI]
ModelDeepSeek V-Flash

DeepSeek V-Flash

A fast inference model released by DeepSeek, designed for speed- and cost-sensitive scenarios. It supports tool-calling capabilities.

Timeline (last 90 days)

Oct 8

In a terminal test against DeepSeek V-Flash, Pi passed 66 out of 89 tasks and DSH passed 65.

Unverified50%
Aug 22

Qwen3 27B is in the same tier as the DeepSeek V-Flash series on the overall intelligence leaderboard, with scores in the 52–53 range.

Expired50%
Aug 22

Qwen3 27B是原生多模态模型,支持视觉输入,弥补了DeepSeek V-Flash无法处理图像的短板

Unverified50%

All Facts (2)

Source Articles