跳到主要内容
AI 视频模型
Kling AI
Kling AI Avatar

Kling AI Avatar Pro

用人物图配合已有音频试做口播,先以品牌试样或课程片段检查口型、动作与画面,再决定是否用于交付。

准备素材
1 张人物图片+1 段目标音频
本站预扣系数
每音频秒 16 积分,合计向上取整
可选提示词
最多 5,000 个字符
输入音频时长
最长 5 分钟,提交前需完成时长检查
快速了解
  • 人物图片和已有音频共同输入,先检查素材内容,再安排试做。
  • 本站按核实音频秒数估算预扣,最终结算还受成功状态与有效用量限制。
  • 把品牌试样和课程对照作为检查方法,关注实际规格、口型与动作。
  • 提示词可以留空;填写时可简要描述希望的人物表现,结果仍需检查。

相关查询: Kling AI Avatar Pro 怎么用 · Kling AI Avatar Pro 预扣费用 · Kling AI Avatar Pro 需要什么素材

示例

Fashion portrait

Cinematic scene

Use the uploaded storyboard as the full visual reference. Follow panels left to right, top to bottom. Generate one continuous 15-second cinematic anime video. Match the storyboard’s character design, clothing, framing, lighting, emotion, and scene order. IMPORTANT: Visuals only. No dialogue, no narration, no vocals, no music, no sound effects, no generated audio. Style: nostalgic late 90s / early 2000s anime film, hand-drawn feeling, painted night backgrounds, soft moonlight, warm yellow-green firefly glow, peaceful magical realism, gentle cinematic motion. 0:00-0:02 — Wide backyard at night under a starry sky. Cool moonlight on grass. A distant forest glows softly with many fireflies. The young boy stands small in frame and notices one tiny light near the ground. Slow push-in. 0:02-0:04 — Medium shot. The boy kneels beside one lone firefly glowing near the grass. Its warm light softly illuminates his curious face from below. Slow dolly down to his eye level. 0:04-0:06 — Close-up of the boy’s hands gently cupping the firefly. Soft glow leaks through his fingers. Cut to his calm smile and wonder-filled eyes. Intimate slow push-in. 0:06-0:08 — Wide-to-medium tracking shot. The boy walks across the moonlit meadow toward the glowing forest, holding his cupped hands near his chest. Tall grass moves gently around him. Warm light pulses softly from his hands. 0:08-0:10 — At the forest edge, the boy stops and slowly opens his hands. The firefly rests on his palm for a quiet beat. Dark trees behind him, many fireflies floating deeper in the forest. Slow push-in. 0:10-0:12 — Close-up as the firefly lifts from his palm, leaving a soft glowing trail. Camera follows it upward, then gently pans as it flies toward the other fireflies. The boy watches in silhouette. 0:12-0:14 — Wide magical forest shot. The lone firefly joins the group, and all fireflies brighten together in one soft golden-green pulse. The forest fills with warm light. The boy stands still, peaceful and small. 0:14-0:15 — Final wide shot. The boy walks back toward his distant home. The glowing forest remains behind him. Stars and moonlight create a calm, nostalgic ending. Slow pull-back. Camera: slow dolly, smooth tracking, stable cinematic movement, soft transitions. Avoid: fast cuts, shaky camera, character morphing, inconsistent face, changed outfit, changed hairstyle, extra limbs, bad anatomy, harsh lighting, flicker, teleporting objects, warped backgrounds, text, watermark, UI.

Fashion portrait

Cinematic scene

Create a 15-second cinematic comedy short in a polished 2D-and-3D hybrid animation style, 16:9 widescreen. Elegant modern restaurant bathroom, beautiful realistic details, expressive character animation, sophisticated visual humor, exaggerated reactions, precise comedic timing, minimal or no dialogue. MAIN CHARACTER: A sophisticated, well-dressed man who takes himself slightly too seriously. His expressions should become increasingly frustrated while he tries to maintain dignity. 0–3 seconds: The man enters an elegant restaurant bathroom and washes his hands. He places one hand underneath an automatic soap dispenser. Nothing happens. He looks at his hand. He tries again. Nothing. 3–6 seconds: He moves his hand closer to the sensor, then farther away. He tries his fingertips, then his palm. He waves his hand underneath the dispenser. Nothing happens. His expression becomes increasingly offended, as if the dispenser has personally rejected him. 6–9 seconds: He begins carefully experimenting with different hand positions. He crouches slightly and studies the sensor. He slowly moves his hand underneath it from different angles. Still nothing. Treat his tiny problem with exaggerated seriousness while keeping the environment completely normal. 9–11 seconds: Another customer casually walks past. Without even looking at the dispenser, she places one hand underneath it. SQUIRT. Perfect amount of soap. She casually walks away. The man watches her. He slowly looks back at the dispenser. His determination becomes extreme. 11–13 seconds: He places BOTH hands underneath the sensor. Suddenly the dispenser releases a huge amount of soap onto his hands. He freezes. Looks at his hands covered in far too much soap. Then looks at the dispenser. A tiny victorious smile appears. He has finally defeated it. 13–15 seconds: He moves to the automatic faucet. Places his hands underneath. Nothing. He tries again. Nothing. He looks at his hands covered in soap. Looks at the faucet. Then slowly looks directly toward the camera with complete defeated disbelief. CUT TO BLACK. STYLE: High-end original 2D/3D hybrid animation, expressive facial acting, elegant cinematic environment, strong visual storytelling, dry observational comedy, exaggerated but believable physical acting, precise timing, escalating frustration. Keep the situation completely ordinary and recognizable. The humor comes from the man's increasingly serious attempts to solve a tiny everyday inconvenience. No surreal events, no explosions, no random chaos, no violence, no complicated dialogue, no confusing camera angles. Make the soap dispenser and faucet clearly visible throughout.

Cinematic scene

Cinematic scene

一句话了解

这个模型怎么样?

Pro 接收人物图片与目标音频,用于尝试说话头像视频。先准备授权素材并完成本站检查;是否适合你的角色、语言和节奏,应以生成结果为准。

模型定位
围绕固定人物图和已有配音试做口播。请先准备好目标音频,再考虑是否用提示词补充希望的人物表现。
适合谁用
适合愿意逐次检查品牌人物或课程讲解效果的创作者;先从便于检查的短片段开始。
本站使用说明
提交前可查看按音频时长计算的预扣估算;完成后结合实际结果和任务明细,核对这次试做是否符合用途。

关键参数一览

根据这些信息,判断模型是否适合你的任务。

模型类别
AI 视频模型
厂商
Kling AI
模型家族
Kling AI Avatar
运行方式
异步任务
图片要求
JPEG 或 PNG,不超过 10 MiB
站内音频上传
不超过 95 MiB,目标音频不超过 5 分钟
成片检查
画面规格、人物呈现、语言、口型与动作以实际结果为准;先检查短片段,再安排交付。
费用核对
提交前看预扣估算;成功后的结算与退差还需确认有效状态和用量,异常费用继续待核。

它还叫什么

同一个模型在公开资料中可能有不同名称,这里整理相关称呼,方便查找和比较。

Kling AI Avatar Pro

用户常见问题

从任务、输入条件和结果要求出发,整理选择模型时需要确认的问题。

Kling AI Avatar Pro 怎么用
Kling AI Avatar Pro 预扣费用
Kling AI Avatar Pro 需要什么素材
Kling AI Avatar Pro prompt 可以留空吗
Kling AI Avatar Pro 音频最长多久
Kling AI Avatar Pro 数字人口播
Kling AI Avatar Pro 和 Standard 怎么选
Kling AI Avatar Pro 和 Infinitalk 区别
Kling AI Avatar Pro 适合品牌口播吗
Kling AI Avatar Pro 与 Standard 预扣费用

选型指南

根据输入素材、任务要求和积分成本,逐项确认模型是否适合。

1
先制作品牌试样

选择有使用权的固定人物图,搭配已确认内容的品牌音频;检查人物呈现、口型和动作是否适合品牌表达。

2
对照课程片段

将同一课程音频分别用于 Pro 与 Standard,逐项比较实际规格、可见缺陷与费用。分别受理的试做属于独立任务,应分别核对明细。

3
长内容先选片段

目标音频不超过 5 分钟;长讲解可先拆成片段试做,检查衔接后再安排完整交付。

模型对比

把常见候选放在同一任务下比较,帮助你确认输入方式、控制项和成本差异。

同一课程素材用 Pro 还是 Standard?
Pro 与 Standard 的本站预扣系数有什么区别?
Infinitalk 与 Pro 对素材和提示词的要求有什么不同?
比较 Pro 与 Infinitalk 时,分辨率和 seed 选项有何不同?

模型横向对比表

按同一组维度比较当前模型和候选模型的输入、设置与使用边界。

素材与提示词
Kling AI Avatar Pro
1 张图片+1 段音频,提示词可留空
Kling AI Avatar Standard
同样使用 1 张图片+1 段音频,提示词可留空
Infinitalk
1 张图片+1 段音频,提示词必填
时长与画面规格
Kling AI Avatar Pro
目标音频最长 5 分钟;成片时长与清晰度需检查实际结果
Kling AI Avatar Standard
目标音频最长 5 分钟;同样需要检查实际成片规格
Infinitalk
可选择 480p 或 720p 请求分辨率
本站预扣计算
Kling AI Avatar Pro
每音频秒 16 积分,合计向上取整
Kling AI Avatar Standard
每音频秒 8 积分,合计向上取整
Infinitalk
按核实音频秒数与所选分辨率估算预扣
可调整的参数
Kling AI Avatar Pro
提示词可选;没有分辨率或 seed 调整项
Kling AI Avatar Standard
提示词也可留空,没有分辨率或 seed 调整项
Infinitalk
提示词必填,提供分辨率与 seed 选项

适用情况

依据公开资料与当前站内设置,说明模型适用的任务及需要注意的限制。

先确定要检查的问题

品牌样本关注人物呈现是否合适,课程样本关注语句与口型是否对应。记录具体问题,再决定重做、换素材或比较其他模型。

比较结果,也核对费用

Pro 与 Standard 可用课程素材作对照,但输出差异和最终费用应分别看实际结果与任务明细,不能从型号名称判断哪种更适合。

能力详解

图片与音频共同输入

使用固定人物图与已准备的配音尝试口播;需要说出的内容应先在目标音频中确认。

可选提示词

提示词可以留空;填写时可简要描述希望人物呈现的表情与动作,不保证每项描述都会实现。

按输入和参数需求选择对照

Standard 也使用图片和目标音频;Infinitalk 要求填写提示词,并提供分辨率与 seed 请求选项。可据此选择需要进一步了解的对照对象。

适合做什么

品牌口播试样

用已获使用权的人物图与品牌音频试做,检查人物呈现、口型和动作,再判断是否适合对外展示。

课程讲解对照

用同一课程音频分别试做 Pro 与 Standard,记录实际画面规格、可见缺陷和各次任务费用,形成自己的选择依据。

短口播交付测试

说明或培训的目标音频不超过 5 分钟;超出时先拆分。片段间是否连贯、是否达到交付要求,仍需人工检查。

提示词技巧

先确认内容,再写提示词

配音中的品牌名称、课程术语与停顿先在音频里确认;提示词可补充希望的人物表现,不能代替需要说出的内容。

对照时保持素材一致

与 Standard 对照时使用相同图片和课程音频,分别保存提示词与检查结果,区分素材变化和模型变化带来的差别。

先用短片段检查人物表现

观察口型与动作是否符合意图,再调整简短描述或换素材重试。另一次试做也要单独核对任务与费用。

为什么选它

可复用已有品牌音频与人物图设计试样,围绕固定表达检查实际结果。
同一课程素材可以分别试做 Pro 与 Standard,把观察到的差异和各次任务费用记录下来。
提示词可留空,先围绕图片和音频做短口播测试,再决定是否补充人物表现描述。

使用前要知道

成片规格、人物呈现、语言、口型与动作都需要检查,不能凭 Pro 名称认定优于 Standard。
上传通过不代表一定能提交或生成:素材还需通过类型、时长等检查,生成结果也需验收。
输入音频不超过 5 分钟;较长内容需要拆分,拆分后也不能保证片段衔接或最终片长。

可调参数

根据这些信息,判断模型是否适合你的任务。

当前页面未列出额外可调设置,请以工作台为准。

积分消耗

本站预扣:16 积分 / 音频秒

提交前按核实的音频秒数估算,以每秒 16 积分相乘后将合计向上取整,音频须不超过 5 分钟。任务受理后按该估算预扣;任务成功、最终状态和用量均有效,且按已确认用量计算的积分向上取整后不超过预扣时,才结算并退还差额。用量缺失、无效、超过预扣或状态待核时,继续保留预扣待核,不追加扣费;失败后的费用与退款请查看任务明细。

使用建议

准备多次生成时,建议先用少量素材测试,核对输入设置和结果是否适合任务,再决定是否继续。每次生成均按工作台显示的计费规则扣点。

常见问题

相关模型

按你的任务比较这些模型的输入要求、功能和当前状态,再查看对应详情。

Kling AI Avatar Standard

用 1 张图片和 1 段目标音频尝试制作说话头像视频,提示词可以留空。可先用短音频检查口型、停顿和人物表现,再决定是否采用结果。

AI 视频模型
每音频秒预扣 8 积分,合计向上取整

Infinitalk

先把要说的话整理成短音频,再配人物图片与非空提示词试做口播。检查人物、语言、口型、动作与构图后,再决定是否用于课程说明、客服答复或虚拟主播内容。

AI 视频模型
本站预扣:480p 每音频秒 3 积分;720p 每音频秒 12 积分

Seedance 2.5

用 Seedance 2.5 规划较长镜头,先选择文字、首帧与首尾帧或多模态参考场景。请求时长可设为最长 30 秒;默认 5 秒,选择自动时长则按 30 秒预留预计积分。准备素材时,数量、单段时长与累计时长需分别核对。

AI 视频模型
预计预扣 102–5,700 积分

信息来源

页面内容依据公开模型资料与 Kyeo AI 的模型信息整理。能否在线使用,请以当前模型状态为准。

来源说明

本页按本站使用要求整理素材要求与预扣说明;成片规格和人物表现仍需结合实际结果检查。

内容复核日期: 2026-08-28
Kling AI Avatar Pro 接口资料(归档)
平台

用于核对本型号的图片、音频输入与提示词说明;本站上传上限和费用规则以本站要求为准,归档资料不保证实际成片效果。