DeepSeek V4-Flash Completes Benchmarks at 1/105th the Cost of Fable

zainhas · x · 2026-08-02

User zainhas shared benchmark results for the DeepSeek V4-Flash model. The data shows that it completes the same benchmark tasks as Fable at 105x lower cost, highlighting the rapidly decreasing expense of LLM intelligence.

Related event: DeepSeek V4-Flash Cuts Costs 100x, Sparking AI Economics Debate(11 posts)→

Original post →

More from Models

Models channel →