智商指数
88.6 #24
AAQI 88.0
输出速度
25 t/s
ITL 40.0 ms/tok
混合价格 (3:1)
$0.1 /1M
输入 $0.1/M · 输出 $0.1/M
上下文
64K
≈ 160 页 A4 · 最大输出 2,048
Quality Domain Radar
6-Axis Frontier Evaluation vs Category Average & Flagship
Kling 3.0 Pro
Claude Mythos Preview (#1)
Average (68)
缺少直接基准测量时,雷达图各轴由类别得分近似得出 — 详见 评分方法。
优势与取舍
优势所在
- Beats Veo/Seedance-class rivals in blind professional evals
- Strong physics — cloth, fluids, collisions stay coherent
- Excellent image-to-video identity retention
需要注意的权衡
- No native audio track — pair with a TTS/score pipeline
- Queue latency on long 1080p generations
- English prompt adherence weaker than Chinese
杀手级用例
UGC ad clips
Short-form product and lifestyle ads at scale.
Previz & storyboards
Animated storyboards with consistent characters.
Product demos
Motion shots from a single reference image.
开发者快速上手与 API
请先设置
KLING_API_KEY — 获取 API 密钥curl -X POST "https://api.klingai.com/v1/videos/text2video" \
-H "Authorization: Bearer $KLING_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model_name": "kling-v3-0-pro",
"prompt": "A drone shot gliding over a rainy neon city at night",
"duration": "5",
"aspect_ratio": "16:9"
}'import os, requests
resp = requests.post(
"https://api.klingai.com/v1/videos/text2video",
headers={"Authorization": f"Bearer {os.environ['KLING_API_KEY']}"},
json={
"model_name": "kling-v3-0-pro",
"prompt": "A drone shot gliding over a rainy neon city at night",
"duration": "5",
"aspect_ratio": "16:9",
},
)
print(resp.json()["data"]["task_id"]) # poll until doneconst res = await fetch("https://api.klingai.com/v1/videos/text2video", {
method: "POST",
headers: { Authorization: `Bearer ${process.env.KLING_API_KEY}`, "Content-Type": "application/json" },
body: JSON.stringify({
model_name: "kling-v3-0-pro",
prompt: "A drone shot gliding over a rainy neon city at night",
duration: "5",
aspect_ratio: "16:9",
}),
});
console.log((await res.json()).data.task_id); // poll until doneAsync video API — submit a task, then poll the task_id until the clip is ready.
官方资源与链接
API 部署
对比托管商与延迟
Groq、Cerebras、Together、DeepInfra、Fireworks、Bedrock 与 Azure 的吞吐量、TTFT 和缓存定价。
GroqCerebrasTogether AIDeepInfraFireworks AIAWS BedrockAzure AI
各榜单得分
分类能力得分
基准测试分项
实时规格与架构
底层解析
Kling 3.0 — diffusion-transformer video model
- Physical-simulation consistency across frames
- Multi-angle camera coherence and motion control
- Pro tier: 1080p output with multi-shot control
- Text-to-video and image-to-video pipelines
发布日期
2026年10月1日
架构
Kling 3.0 — diffusion-transformer video model
参数量
Not disclosed
开源协议
Proprietary
上下文窗口
64,000 tokens
最大输出
2,048 tokens
模态
text + video
缓存价格
读 $0.02 · 写 $0.1 /1M
推理
—
输入价格
$0.1/M
输出价格
$0.1/M
输出速度
25 t/s
首Token延迟
—
数据更新
2026年10月5日
每月账单预估
每月预估支出
$0.00
正在计算相对 GPT-4o 的节省额…
引用与嵌入
GitHub Markdown 徽章
Loading badge markdown...BibTeX
@misc{llmpodium2026kling30pro,
title = {Benchmark Analysis and Intelligence Evaluation of Kling 3.0 Pro},
author = {LLMPodium Research},
year = {2026},
url = {https://llmpodium.com/models/kling-3-0-pro}
}嵌入评分徽章
将实时 Podium 评分徽章添加到 GitHub 或 Hugging Face README —— 链接回实时排名。
预览
Markdown
Loading badge snippet...HTML
Loading badge snippet...更多 Kwai 模型
替代模型推荐
基于能力、定价和基准测试得分与 Kling 3.0 Pro 最接近的替代模型。