DeepSeek model lineup
Every name here is copied from DeepSeek’s own spelling. Short blurbs are mine, after reading the launch notes and the pricing table.
DeepSeek-V4-Pro
Flagship V4 model: 1.6T total / 49B active, 1M context, Thinking and non-Thinking, built for agents and hard coding.
DeepSeek-V4-Flash
Faster V4 model: 284B total / 13B active, 1M context, Instant Mode on chat, economical API default.
DeepSeek-V4-Flash-Vision-Exp
Experimental V4-Flash that accepts image input (JPEG, PNG, GIF, WebP) alongside text on the API.
DeepSeek-V3.2
Reasoning-first predecessor: Thinking in tool-use, live on App, Web, and API when it shipped.
DeepSeek-V3.1
Hybrid Think / Non-Think generation, 128K context at launch, first step toward the agent era.
DeepSeek-R1
Reasoning model trained with large-scale RL; MIT weights; DeepThink on chat when it launched.
DeepSeek-V3
671B MoE / 37B active open model that powered the free app at launch.
Need a job instead of a version? Try agentic coding or the API platform.