DeepSeek V3 0324

DeepSeek released 2025-03-24

$0.438 per million tokens, balanced

DeepSeek V4 Flash 0731 is both better and cheaper.

It scores +36.6 higher and costs 76% less ($0.105/M against $0.438/M) on this workload — and it does everything this model does.

Solar Pro 4 is cheaper still (88% less) but drops 164K → 131K max output.

Every price dimension

Input$0.250/M
Output$1.00/M

Independent scores

Intelligence15.2
Coding21.2
Agentic1.6
LMArena Elo1375default

Sources are listed separately rather than averaged. Across the 62 models both have scored they correlate at r = 0.806 — close agreement overall, but four models rank very differently between them, and a blended score would hide exactly those.

Capability

Context window164K
Max output164K
Input modestext
Tool useyes
Reasoningno
Knowledge cutoff2024-07-31

Provenance

Price sourceopenrouter.ai
Fetched2026-08-24
Quality dataverified

DeepSeek V3, a 685B-parameter, mixture-of-experts model, is the latest iteration of the flagship chat model family from the DeepSeek team. It succeeds the [DeepSeek V3](/deepseek/deepseek-chat-v3) model and performs really well...

Markdown for LLMs