The AI tools index that doesn't waste your time.不浪费你时间的 AI 工具索引。
Head-to-head正面对比

Kimi K3 vs DeepSeek V4 Pro 0423

Moonshot AI meets DeepSeek: sourced benchmark scores side by side, dimension by dimension, with prices and specs. Every conclusion below is derived from the published dataset — no opinions, no sponsorship.

Moonshot AI 对 DeepSeek:带来源的 benchmark 成绩逐项并排,附价格与规格。下面每条结论都从公开数据集推导——没有主观意见,没有厂商赞助。

◈

The verdict, derived from data结论(全部由数据推导)

ReaCodKnoMatAgtPre

● Kimi K3   ● DeepSeek V4 Pro 0423 · missing dimensions count as 0 in the chart缺数据的维度在图上按 0 计

≡

Spec by spec逐项规格

Kimi K3DeepSeek V4 Pro 0423
Provider厂商Moonshot AIDeepSeek
Released发布2026-07-162026-04-24
Intelligence Score智能评分87.987.4
Benchmark coveragebenchmark 覆盖95%95%
Reasoning推理87.589.2
Coding代码97.785.9
Knowledge知识9594.5
Math数学70.785.5
Agent智能体93.897.1
Preference偏好66.558.8
API input $/1M输入价 $/百万$0.300$0.372
API output $/1M输出价 $/百万$14$0.743
Blended $/1M混合价 $/百万$7.15$0.558
Value (score per $)性价比(分数/美元)12.3156.8
Context window上下文1M1M
Max output最大输出944K384K
Reasoning tiers推理档位always-on thinkingthinking off / high effort / max effort
Open weights开放权重No否Yes支持
CN-direct国内直连Yes支持Yes支持
Free tier免费层No否No否
⚔

Benchmark by benchmark逐项 benchmark

Benchmark基准Kimi K3DeepSeek V4 Pro 0423
Humanity's Last Exam44.348.2⚡max
GPQA Diamond93.590.1
MMLU-Pro8887.5
SWE-bench Verified93.480.6
AIME 202596.296.67⚡max
LiveCodeBench93.593.5⚡max
FrontierMath39⚡max68.3
Terminal-Bench88.367.9⚡max
τ²-bench9396.2
LMArena (Chatbot Arena)1489⚡max1458

Kimi K3 dossier → 档案 → · DeepSeek V4 Pro 0423 dossier → 档案 → · How we score评分方法