跳到主要内容
AI 视频模型
Kling AI
Kling

Kling V2.1 Master

从文字开始构思视频,或上传 1 张首帧图,再描述希望出现的动作。两种输入方式都按时长预计预扣:5 秒 160 积分,10 秒 320 积分。

任务类型
文生视频、图生视频
预计预扣
5 秒 160;10 秒 320 积分
提示词上限
5,000 字符
上传限制
图生:JPEG/PNG 1 张,最大 10 MB
快速了解
  • 支持文字描述或单张首帧图两种输入
  • 图生需 1 张 JPEG/PNG 图片,最大 10 MB
  • 5 秒预计预扣 160 积分;10 秒预计预扣 320 积分
  • 文字生成可选画面比例;图生无画幅设置

相关查询: Kling V2.1 Master 怎么用 · Kling V2.1 Master 价格 · Kling V2.1 Master 教程

示例

Cinematic scene

CAMERA: DV 16mm tape camcorder, handheld POV/self-shot by CHASE. Occasional propped shots on bench/floor. Shaky framing, soft focus, clumsy zooms, face cutoffs, chaotic movement during exercise, steadier during recovery. Camcorder never visible. LOOK: Soft blurry tape quality, faint noise, bloomed gym lights, flickering exposure, muted contrast, realistic skin tones. Sweat progressively builds. STYLE: High-energy, breathless HIIT vlog. Intense exercise bursts → short recovery beats. Genuine exhaustion and determination. CHASE: Korean idol in her 20s, long black high ponytail becoming damp, expressive eyes, flushed cheeks, slim athletic build. Modest long-sleeve athletic top, joggers/leggings, sneakers, towel, no jewelry. SETTING: Evening gym open floor, mats, mirrors, small speaker/phone, water bottle, soft overhead lighting. 30s / 9 CUTS: 1. 3s Propped: Ready on mat, bouncing. “Okay, HIIT round thirty seconds on, let’s go!” 2. 3s Handheld: Fast jumping jacks. “Okayalready feeling this” 3. 2.5s Propped: Hands on knees, recovering. “Okay five seconds that’s it” 4. 3s Low/propped: Intense mountain climbers, sweat visible. Heavy breathing only. 5. 2.5s Propped: Sitting back, wiping sweat. “Okay, that one’s brutal” 6. 3.5s Handheld: Burpees, increasing motion blur. “Why did Ipick burpees” 7. 2.5s Macro: Hands gripping mat, sweat dripping, chest heaving. No dialogue. 8. 3s Propped: Collapses onto back, laughing breathlessly. “Okay I think I’m dying” 9. 3.5s Selfie: Slowly sits up, drinks water, tired grin. “Okay, that’s a wrap worth it though. See you guys!”

Cinematic scene

Cinematic scene

Create a completely original, spontaneous personal video of a young Korean woman during a warm summer afternoon in Seoul. It should feel like her friend happened to record a completely ordinary errand that unexpectedly turned into a memorable little moment. MAIN CHARACTER Young Korean woman in her early 20s, naturally attractive, realistic pores and skin texture, minimal makeup, natural blinking and subtle expressions. Long straight black hair loosely clipped at the back with a few strands falling naturally around her face. WARDROBE: loose white cotton button-up shirt with rolled sleeves, muted green pleated midi skirt, simple cream sneakers, small black shoulder bag and a thin silver bracelet. Keep her face, identity, hairstyle, proportions, clothing and accessories completely consistent. LOCATION A lived-in Seoul residential neighborhood during a hot summer afternoon. Narrow sloping street, older brick houses, small vegetable gardens, metal gates, parked bicycles, utility poles, overhead cables, apartment buildings visible in the distance and a tiny neighborhood laundromat. CAMERA / REALISM The entire video is casually filmed by a friend on an older compact digital camera. Natural handheld movement, imperfect framing, slight camera shake while walking, occasional autofocus hunting, small exposure shifts, mild motion blur, subtle digital noise and authentic consumer-camera compression. No cinematic stabilization, drone shots, dramatic lighting, beauty filters or polished commercial cinematography. --- 00:00–00:05 — QUICK ERRAND She steps out of her house carrying a small reusable tote. She checks a handwritten shopping note, folds it and puts it into her pocket. She starts walking down the street while casually talking to her friend behind the camera. A neighborhood delivery scooter passes slowly, and she naturally steps aside to let it through. 00:05–00:10 — THE LAUNDRY SHOP She enters a tiny self-service laundromat. She takes a freshly dried white shirt from a folding table and tries to fold it while walking. The shirt slips from her hands and falls onto the floor. She looks at it for a second, laughs at herself and picks it up. The camera operator chuckles. 00:10–00:15 — SMALL COINCIDENCE She walks back outside. A sudden breeze blows several lightweight receipts and paper slips across the sidewalk. One gets stuck against her shoe. She notices it, picks it up and looks around for where it came from. A nearby shopkeeper gestures toward a small recycling box. She walks over and drops it inside. 00:15–00:20 — UNEXPECTED LITTLE MOMENT As she turns around, she notices a small handwritten “free vegetables” basket outside a neighbor's gate. Fresh tomatoes and green peppers sit inside. She looks toward the house, calls out politely, and takes one tomato after receiving a friendly gesture from inside. She places it carefully into her tote. 00:20–00:25 — THE REALISTIC DETAIL She continues walking home. The tomato rolls out of the loosely closed tote and drops onto the sidewalk. She quickly catches it with her foot before it reaches the street. She looks at the camera in disbelief, then laughs. She fixes the bag properly this time. 00:25–00:30 — NATURAL ENDING She reaches her gate. Before going inside, she looks at the camera and holds up the tomato. With a small laugh she says: “At least I saved this one.” She opens the gate and walks inside. The friend keeps filming for another second as the gate closes naturally. PHYSICAL REALISM Every action must obey real-world physics. The shirt must actually fall, paper must react naturally to wind, the tomato must roll realistically and stop when her shoe catches it, and the tote must behave like a lightweight fabric bag. Natural hand anatomy, fingers, feet, facial expressions, hair movement and clothing physics.

Cinematic scene

Visual transformation

Cinematic scene

Create a 60-second ultra-realistic cinematic wedding film featuring an adult bride and adult groom, both around 25–30 years old, with elegant, natural, photorealistic faces and consistent identity throughout the entire video. The wedding takes place in an extraordinary dreamlike location unlike a typical wedding venue: a magnificent glass cathedral garden built beside a crystal-clear lake, surrounded by tall white flowering trees, floating candle lanterns, soft mist over the water, thousands of tiny warm lights, white roses, ivory orchids, and elegant hanging crystal decorations. The setting feels luxurious, magical and believable, never fantasy-looking or artificial. Golden-hour sunlight slowly transitions into a warm romantic evening glow. Wedding fashion: The bride wears a breathtaking luxury white Christian bridal gown, floor-length with an elegant flowing train, delicate lace details, long soft veil, subtle diamond earrings, natural bridal makeup, perfectly styled hair and realistic fabric movement. The groom wears a perfectly tailored deep black tuxedo, crisp white shirt, black bow tie, polished black shoes and a small white rose boutonnière. All wedding guests are stylish adult men and women wearing a coordinated black, white, champagne and soft beige dress theme, creating an exclusive high-fashion wedding atmosphere. 0:00–0:08 — Grand Opening: Begin with a cinematic aerial shot moving above the lake toward the glass wedding cathedral. Soft church bells blend into emotional piano and cinematic strings. The camera enters through the flower-covered entrance where guests stand gracefully on both sides. The bride slowly walks toward the groom while her long white veil moves naturally in the breeze. Use close-ups of her smile, the groom’s emotional expression, flowers and wedding details. 0:08–0:17 — Meeting & Bouquet Moment: The groom walks toward the bride through a pathway covered with white flower petals. He holds an elegant bouquet of white roses, orchids and small green leaves. He gently kneels on one knee in front of her and presents the flowers with a warm smile. The bride looks surprised and emotional, then accepts the bouquet. Keep body movement natural and realistic. Groom softly says: "You are the most beautiful chapter of my life, and today I choose you for every tomorrow." Bride smiles and replies: "And wherever life takes us, I want every road to begin with you." 0:17–0:30 — Walking & Romantic Wedding Shots: They walk hand-in-hand beside the lake through the illuminated garden while guests smile in the background. Show cinematic tracking shots from the front, side and behind. The bride’s gown and veil flow beautifully while the groom occasionally looks toward her. Include close-ups of their hands, bouquet, wedding rings, smiling faces, reflections in the water and glowing lanterns. 0:30–0:43 — Elegant First Dance: The music rises into a beautiful emotional orchestral wedding theme with soft piano, strings and subtle choir ambience. The groom gently invites the bride into a slow first dance beneath a large circular floral installation filled with warm fairy lights. He guides her into a graceful turn; her white gown creates a beautiful circular motion around them. The camera smoothly orbits around the couple while they continue a natural, elegant slow dance. Guests watch respectfully from a distance. During the dance, the groom quietly says: "I wish this moment could last forever." The bride replies with a soft smile: "Then let’s remember it forever." 0:43–0:53 — Ceremony & Vow Moment: Transition to the couple standing beneath a white floral wedding arch inside the glass cathedral. An adult officiant stands beside them while guests remain seated. Soft candles and sunlight illuminate the scene. The couple exchange rings and look into each other’s eyes. The groom says: "From this day forward, I promise to walk beside you through every season of life." The bride replies: "And I promise to make every season worth remembering." 0:53–1:00 — Wedding Kiss & Final Cinematic Ending: After the vows, the couple share a brief, tasteful wedding kiss while the guests applaud and white flower petals gently fall around them. The orchestral music reaches its emotional peak. Immediately after, they smile, hold hands and slowly walk beneath the flower arch as the camera pulls backward. Finish with a dramatic wide aerial shot showing the illuminated glass cathedral, lake, floating lanterns and wedding guests as evening begins. End on the bride and groom looking at each other and smiling naturally. Visual style: ultra-realistic live-action wedding film, premium luxury wedding cinematography, natural skin texture, realistic eyes and facial expressions, accurate hands, consistent faces, physically realistic fabric, cinematic depth of field, soft golden-hour lighting, elegant camera movement, subtle handheld moments mixed with smooth gimbal and drone shots, realistic flower movement, professional wedding color grading, 4K detail, realistic shadows and reflections. Audio: emotional cinematic wedding score with piano, violin, soft strings, gentle choir atmosphere, occasional church bells, subtle natural ambience, guests softly applauding, footsteps, fabric movement and realistic environmental sounds. Dialogue must remain clear, emotional and naturally spoken without sounding robotic. Important continuity: Keep the exact same adult bride and groom throughout all shots. Do not change faces, hairstyles, clothing, body proportions, bouquet, wedding location or wedding theme. No cartoon appearance, no artificial AI faces, no exaggerated expressions, no distorted hands, no random costume changes, no sudden scene jumps. Maintain smooth cinematic transitions and a believable luxury Christian wedding atmosphere for the full 60 seconds.

一句话了解

这个模型怎么样?

文字生成可选择画面比例;图生使用单张首帧图,不提供画幅设置,也不能添加尾帧图。两种输入都需要提示词,并可调整时长、反向提示词和提示词贴合度。生成后请检查画面和动作,再决定是否使用。

模型定位
Master 属于 Kling AI 2.1 系列。在 Kyeo AI,可用文字描述生成视频,也可从单张首帧图开始。当前表单没有独立分辨率选项。
适合谁用
可用于短分镜草样,或以商品图、角色图为起点尝试动作和运镜。请检查主体细节、文字和动作是否符合要求。
本站使用说明
先按手头素材选择文字生成或图生,再查看所选时长的预计积分。任务提交后,可在作品中查看进度、结果和积分处理提示。

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
品牌
Kling AI
模型家族
Kling
输入方式
文字或单张首帧图
图片输入
1 张 JPEG/PNG 首帧图
画面比例
文字生成有 3 种选项;图生不提供画幅设置
运行方式
提交后查看任务进度和结果
版本名称
Master
提示词上限
5,000 字符
上传限制
图生需 JPEG/PNG 1 张,最大 10 MB

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 2.1 Master
Kling V2.1 Master
Kling 2.1 Master 输出分辨率
Kling 2.1 Master 视频时长

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 V2.1 Master 怎么用
Kling V2.1 Master 价格
Kling V2.1 Master 教程
Kling V2.1 Master 适合哪些视频任务
Kling V2.1 Master AI 视频生成
Kling V2.1 Master 文生视频
Kling V2.1 Master 图生视频
Kling V2.1 Master 比例选项
Kling V2.1 Master 图片输入
Kling V2.1 Master 画幅设置
Kling V2.1 Master 时长
Kling V2.1 Master 视频生成模型

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
手头有文字构思或首帧图

只有文字构思时,先写清主体、场景和动作;已有首帧图时,上传图片并补充运动描述。按任务需要选择输入方式,不必为比较而把两种任务都提交一次。

2
需要尾帧图或更多镜头设置

需要指定尾帧图时,可查看 V2.1 Pro 或 V2.5 Turbo Pro;需要单独编排多镜头时,可查看 Kling 3.0 的设置。提交前分别核对这些模型的当前输入要求和预计积分。

3
先检查一段短视频

可从 5 秒开始,检查构图、主体、动作和画面中的文字。继续尝试时,尽量保持主要输入一致,每次只调整一个设置。每次重新提交的任务被受理后都会预扣积分。

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。

Kling V2.1 Master 和 Kling V2.5 Turbo Pro 有哪些输入差别?
Kling V2.1 Master 和 Kling V2.5 Turbo Pro 怎么选?
Kling V2.1 Master 和 Kling V2.1 Pro 能用哪些图片?
Kling V2.1 Master 和 Kling V2.1 Pro 的预计积分有什么不同?
Kling V2.1 Master 和 Kling 3.0 适合哪些视频任务?
需要多镜头时,Kling V2.1 Master 和 Kling 3.0 怎么选?

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。

任务类型
Kling V2.1 Master
文字生成或首帧图生成
Kling V2.5 Turbo Pro
文字生成或首帧图生成,可选尾帧图
Kling V2.1 Pro
仅图生视频
图片输入
Kling V2.1 Master
图生使用 1 张首帧图,不支持尾帧图
Kling V2.5 Turbo Pro
首帧图与可选尾帧图,最多 2 张
Kling V2.1 Pro
首帧图与可选尾帧图,最多 2 张
画面比例
Kling V2.1 Master
文字生成可选 16:9、9:16、1:1;图生无画幅选项
Kling V2.5 Turbo Pro
文字生成可选 16:9、9:16、1:1;图生无画幅选项
Kling V2.1 Pro
不提供比例控件
预计预扣积分
Kling V2.1 Master
5 秒 160;10 秒 320
Kling V2.5 Turbo Pro
5 秒 42;10 秒 84
Kling V2.1 Pro
5 秒 50;10 秒 100
提示词上限
Kling V2.1 Master
5,000 字符
Kling V2.5 Turbo Pro
2,500 字符
Kling V2.1 Pro
5,000 字符

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。

先决定画面从哪里开始

没有首帧图时,用文字说明初始场景;有图片时,以图中主体为起点补充动作和运镜。图片不能保证主体、构图或文字在视频中保持不变。

比较时长和预计积分

本模型两种输入方式的预计预扣相同,时长决定预扣档位。与其他模型比较时,先核对各自的预计积分,再按同一创作要求检查结果,不用费用高低推断画面好坏。

能力详解

从文字或首帧图生成视频

文字生成时描述主体、场景和镜头;图生时上传 1 张首帧图,再写希望出现的动作。两种方式都可选 5 秒或 10 秒,提示词最多 5,000 字符。

按输入方式选择画幅

文字生成可选 16:9、9:16 或 1:1。图生没有画幅控件,请先准备适合用途的首帧图;结果的比例和构图仍可能变化,生成后需要检查。

调整反向提示词与贴合度

反向提示词可填写希望避开的内容,最多 500 字符。提示词贴合度范围为 0~1,默认 0.5,可按 0.1 的间隔调整;请观察实际结果,不把更高数值当作更好效果的保证。

适合做什么

文字分镜草样

写清主体、动作、场景和镜头,先尝试一段 5 秒视频。查看人物和物体的位置、动作衔接及画面中的文字,找出需要调整的地方。

商品或角色的动作草稿

上传一张有权使用的首帧图,描述主体动作和镜头移动。使用结果前,检查商品结构、角色外观、服装及肢体动作是否符合要求。

比较不同输入方式

如果需要比较文字生成与图生,可尽量保持主要提示词和时长一致,逐项查看构图、光线、主体和动作。记录需要调整的地方,按这段镜头的要求选择结果。

提示词技巧

把动作和镜头写清楚

文字生成时写清主体、场景和动作;使用首帧图时,重点补充动作、镜头方向、速度和环境变化,避免只重复图片里已有的内容。

先选画幅或准备首帧图

文字生成可选 16:9、9:16 或 1:1。图生没有独立画幅设置,上传前请检查图片构图,生成后再核对视频是否适合预定用途。

每次只调整一个设置

继续尝试时,尽量保持输入方式和主要提示词一致,单独调整时长、反向提示词或贴合度,再比较结果。这样更方便找出哪些修改值得保留。

为什么选它

同一模型页面提供文字生成和单张首帧图生成两种输入方式。
文字生成提供 3 种固定画幅;图生使用首帧图,不要求填写画幅参数。
两种输入方式按同一时长档位预计预扣,提交前可查看本次预计积分。
图生需 1 张 JPEG/PNG 图片,最大 10 MB。

使用前要知道

图生只使用 1 张首帧图,不能添加尾帧图或多张参考图。
图生只接受 JPEG 或 PNG,单张不超过 10 MB,不接受 WebP。
本模型没有视频数量、独立分辨率或多镜头编排选项;生成结果不保证主体、文字和构图保持不变。
需要尾帧图或更多镜头设置时,请查看对比模型各自的输入选项、当前状态和预计积分。

可调参数

根据这些信息,判断模型是否适合你的任务。

视频时长
可选
参数类型: 下拉选择
默认 5 秒
5 秒
10 秒
宽高比
仅文字生成
参数类型: 下拉选择
默认 16:9
16:9
9:16
1:1
反向提示词
可选
参数类型: 文本输入
最多 500 字符
提示词贴合度
可选
参数类型: 数值输入
默认 0.5
0~1,步长 0.1

积分消耗

预计预扣 160~320 积分 / 任务

文字生成和首帧图生成的预计预扣相同:5 秒 160 积分,10 秒 320 积分。任务受理后先预扣,生成成功并确认实际消耗后结算;成功但消耗尚未确认时,仍保留预扣。失败后请查看任务的积分退还提示,并到积分明细核对记录。实际恢复的可用积分受原有效期和相关欠款抵扣影响。前置校验或限流未通过时不会预扣;重新提交的任务被受理后会再次预扣。

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 V2.5 Turbo Pro

用文字描述生成视频,或上传首帧图片并按需补充尾帧。Kling V2.5 Turbo Pro 可选 5 秒和 10 秒,两种生成方式使用相同的预计预扣积分档位。

AI 视频模型
预计预扣 42–84 积分

Kling V2.1 Pro

准备首帧图,用提示词描述画面接下来如何变化。需要参考结束画面时,再加入尾帧图;最多 2 张,按首帧、尾帧的顺序上传。选择 5 秒或 10 秒后,先查看预计积分再提交。

AI 视频模型
5 秒预计预扣 50 积分,10 秒 100 积分

Kling 3.0

用文字或图片制作视频,可选择单镜头,也可逐镜头安排提示词与时长。声音和画质档位会参与预计预扣计算;人物、口型、文字、声音及镜头衔接都需检查实际结果。

AI 视频模型
预计预扣 42~1,005 积分/次

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

输入要求、参数和预计积分以 Kyeo AI 当前页面为准;下方官方公告可用于了解 Kling 2.1 系列的发布背景。

内容复核日期: 2026-08-27
快手:Kling AI 一周年与 2.1 系列发布
官方
查看原始来源
快手 2025 年第二季度及中期业绩公告(Kling AI 2.1)
官方
查看原始来源