Skip to main content
vybing.dev
Try:
All models

DeepSeek

DeepSeek V3.1

Frontier-class open-weights MoE model with very low hosted inference cost. Strong coding and math evals.

At a glance

Context
128Ktokens
Input
$0.27/ M tok
Output
$1.1/ M tok
Released
2025-12-26

Vendor