โก DeepSeek V4-Flash is now the cheapest major AI model to run globally
DeepSeek officially released V4-Flash on July 31 โ a 284B-parameter MoE model with only 13B active params and a 1M-token context window, tuned for fast inference. Research firm Artificial Analysis confirmed it costs roughly 3 cents per benchmark test, the least expensive well-known AI model in the world. It scores 50/100 on the Intelligence Index, holding its own on coding and reasoning. Could extreme cost-efficiency finally unlock AI for everyday apps that were priced out before?
โก 22 buzzing
382 views
๐ 0 subscribed