All Models
DeepSeek V4-Flash
284B total, 13B active params; extremely cheap, MIT licensed open-weights model.
DeepSeek V4-Flash is the lightweight counterpart to V4-Pro, using 284B total parameters with only 13B active. At $0.14/$0.28 per million tokens, it is one of the cheapest performant models available from any provider. MIT licensed and open-weights, it can be self-hosted. It is best suited for high-volume text processing pipelines, extraction tasks, and scenarios where cost must be minimized while maintaining reasonable capability.
DeepSeek V4-Flash
Made by
DeepSeek
Released
Apr 24, 2026
Context Window
1M
Pricing
$0.14 / $0.28 per 1M tokens
Provider
DeepSeek