跳至主要內容
AI 影片模型
Alibaba
Wan

Wan 動畫替換

Wan 2.2 Animate Replace 使用工作台上傳的一段來源影片和一張替換圖片完成角色替換,並提供 480p、580p 或 720p 三檔。它按已驗證來源影片秒數計費,使用者不能直接填寫任意媒體 URL。

任務契約
1 段已驗證來源影片 + 1 張替換圖片
目前點數
6/9.5/12.5 credits × 影片秒
提示詞
介面不提供
素材邊界
工作台上傳;目前公開資料介面表單列影片與圖片各 10 MB
30 秒快速瞭解
  • Wan 動畫替換接受哪些影片與圖片格式? — 使用 Wan 2.2 Animate 在 Kyeo AI 以一段上傳影片和一張圖片替換主體;480p/580p/720p 按每個已驗證影片秒計費 6/9.5/12.5 credits。
  • 目前公開資料三檔費率為 480p 每秒 6、580p 每秒 9.5、720p 每秒 12.5 credits
  • 目前公開資料介面表單列影片 MP4/QuickTime/Matroska、圖片 JPEG/PNG/WebP,單檔最大 10 MB
  • 使用來源影片、替換圖片與解析度,送出前可直接核對輸入

熱門搜尋: 如何使用 Wan 動畫替換? · Wan 動畫替換需要上傳哪些素材? · Wan 動畫替換接受哪些影片與圖片格式?

示例

Cinematic scene

Main Subject: Young Japanese woman, early 20s, effortless natural beauty, warm genuine smile, sun-kissed realistic skin texture, minimal makeup, relaxed and friendly personality. Wearing a washed sand-beige oversized linen shirt worn open over a muted seafoam ribbed tank top, loose ivory drawstring linen pants with rolled cuffs, cream canvas sneakers carried in one hand while walking barefoot on the sand, delicate gold necklace, small silver hoop earrings, black straight hair tied in a loose low ponytail with soft face-framing strands. Maintain identical identity, clothing, hairstyle, facial features, and appearance consistently throughout the entire video. Location: Quiet Japanese coastline during a peaceful late afternoon in summer. Wide sandy beach, gentle Pacific waves, weathered concrete seawall, driftwood scattered naturally, distant fishing boats, rocky shoreline, native coastal grasses moving in the wind, a small lifeguard tower in the distance, clear horizon, soft sea breeze. No crowds, no beach bars, no umbrellas, no advertisements, no commercial activity, no staged scenery. Authentic local atmosphere. Visual Style: Ultra-realistic observational documentary. Genuine slice-of-life moments. Completely candid behavior. Natural body language. Rich environmental details. Everyday realism. Warm emotional authenticity. Contemporary 2026 color science with soft natural contrast, realistic skin tones, gentle highlight roll-off, subtle pastel coastal palette, slightly airy exposure, faithful HDR lighting, no stylized filters. Camera Style: Modern compact mirrorless camera recording handheld by a close friend. Natural handheld movement with slight micro-shake. Occasional focus breathing. Face detection occasionally misses focus. Small framing imperfections. Gentle rolling shutter during quick movement. Realistic autofocus transitions. Natural lens characteristics. No cinematic gimbal shots. No dramatic drone footage. No excessive slow motion. No artificial camera moves. Feels like an authentic personal vacation recording from 2026. --- 00:00–00:02 She walks barefoot near the shoreline carrying her sneakers, letting small waves wash over her feet. Wind gently lifts loose strands of hair. She laughs quietly while looking toward the ocean as autofocus briefly searches. 00:02–00:04 She notices several small seashells and crouches to examine them. The camera arrives slightly late and reframes awkwardly while trying to keep her centered. Gentle surf fills the background. 00:04–00:06 She picks up an interesting shell, smiles naturally, then places it back into the sand. A stronger breeze moves her oversized shirt. Sunlight sparkles across the water behind her. 00:06–00:08 She walks along the seawall sipping iced tea from a clear reusable bottle. A cyclist passes in the distant background. The camera briefly overexposes as it pans toward the bright ocean before correcting itself. 00:08–00:10 She sits on a large piece of driftwood watching waves roll in. She absentmindedly brushes hair behind her ear while listening to seabirds. Loose handheld side angle slowly drifts without deliberate composition. 00:10–00:12 Close three-quarter profile. Someone behind the camera jokes with her. She turns unexpectedly, smiles with genuine amusement, softly laughs, and casually says, "Kirei da ne." ("It's beautiful, isn't it?") The camera catches her reaction a moment too late. 00:12–00:15 She continues walking slowly along the water's edge, occasionally glancing toward the camera before returning her attention to the horizon. Small waves reach her feet. The recording ends abruptly mid-step as if the camera operator accidentally stopped recording. Audio: Natural location sound only — gentle ocean waves, distant seabirds, soft coastal wind, footsteps in wet sand, occasional fishing boat engine far offshore, fabric moving in the breeze, subtle water splashes, quiet conversation from far away...

Cinematic scene

A cinematic, hyper-detailed video of an elegant young woman in a dreamlike rose garden bathed in warm, soft golden hour sunlight. She has dark wavy hair styled in a romantic updo adorned with fresh red roses and delicate dangling jewelry, wearing a dramatic sheer black and dark crimson gown embellished with intricate red floral appliqués and ruby-like beads, along with a matching choker and drop earrings. Holding a luxurious, faceted ruby-red perfume bottle with an ornate gold stopper near her face, she slowly turns her head in smooth, graceful slow motion, bringing the perfume bottle across her chest as delicate red rose petals float gently through the air around her while the camera pans softly to capture the dreamy bokeh background of blooming roses.

Cinematic scene

Ceate a premium French pharmacy skincare commercial for La Roche-Posay Anthelios UVMune 400 Hydrating Cream SPF50+. 35mm anamorphic, 2.39:1, ultra-realistic, warm natural film grain, elegant ivory and stone palette with subtle La Roche-Posay orange and clinical blue accents. No lens flares. OPENING — GIRL FIRST: Start immediately with an intimate close-up of the female model’s face in soft morning window light. She looks naturally into the camera with realistic pores, fine facial texture and minimal makeup. She gives a subtle confident smile while holding the sunscreen tube beside her cheek. Her face must be clearly visible in the very first frame and remain consistent throughout. MORNING RITUAL: She stands in a bright minimalist bathroom wearing a loose cream linen shirt with full sleeves. She places the tube beside the sink and picks it up naturally. Quick macro shot of her hands opening the cap and dispensing a small amount of cream onto her fingertips. TEXTURE: Extreme macro shot of the cream being gently spread across the back of her hand. Show realistic creamy texture, soft movement and natural absorption without exaggerated shine. APPLICATION: Cut back to the woman applying the sunscreen evenly across her face while looking into a mirror. Show natural finger movement, realistic skin texture and the cream gradually becoming invisible as it blends in. OUTDOOR TRANSITION: She steps outside into warm morning sunlight, walking through a quiet European street. The camera follows naturally as she adjusts her hair and smiles softly. Her skin remains realistic and naturally textured, never overly glossy. PRODUCT MOMENT: Close-up of the sunscreen tube in her hand as sunlight falls across the packaging. Capture the exact product shape, colors, typography and details without distortion. FINAL FRAME: Return to a close-up of her face. She looks directly into the camera with a relaxed, confident expression while holding the product near her shoulder. The camera slowly moves closer, keeping both her face and the product visible. ON-SCREEN TEXT: “Protection that feels like skin.” “Everyday. Invisible. Comfortable.” MODEL / CONTINUITY: Same female model throughout, warm medium skin tone, dark brown hair in a clean low bun, refined oval face, natural brows, subtle lips, minimal makeup. Same face, hairstyle, wardrobe, skin tone and proportions in every shot. AUDIO: Soft premium skincare music with gentle piano and subtle ambient textures. Add realistic tactile sounds: cap opening, cream dispensing, fingertips rubbing, bathroom ambience, footsteps and light outdoor sounds. Music remains subtle and never overpowering. STYLE: International French pharmacy aesthetic, premium but believable, cinematic macro photography, natural morning light, realistic skin, tactile product details, shallow depth of field, elegant camera movement, warm film grain, photorealistic 4K. NEGATIVE: No waxy skin, beauty filters, excessive retouching, oily shine, distorted product, changing packaging, distorted hands, extra fingers, identity changes, text glitches, exaggerated acting, lens flares, watermark or unrealistic CGI appearance.

Cinematic scene

Main Subject: Young Turkish woman, early 20s, native Anatolian appearance. Naturally beautiful with warm olive-toned skin, realistic skin texture, minimal to no makeup. Curvaceous feminine figure with a full bust and narrow waist, wearing a fitted pale-cream cotton blouse with the top button undone, tucked into a long flowing floral-patterned skirt reaching her ankles, simple flat leather sandals. A light, loosely tied traditional white cotton headscarf (yemeni) with delicate lace edging, worn casually so wisps of dark hair frame her face. Thin gold bracelet on one wrist. Warm, shy, approachable personality with a natural graceful posture. Maintain consistent identity, clothing, hairstyle, and appearance throughout the entire video. Location: Authentic Anatolian village during a calm late morning. Whitewashed stone and mud-brick houses with terracotta roofs, weathered wooden doors, a traditional stone village fountain (çeşme) with water flowing continuously from a carved stone spout into a worn stone basin. Cobblestone and dirt paths, grapevine pergolas casting dappled shade, poplar trees, chickens roaming freely, copper and plastic water jugs, laundry lines, low stone walls, distant fields. Quiet village atmosphere. No modern buildings, no cars, no commercial elements. Visual Style: Ultra-realistic documentary realism. Genuine candid behavior. Natural body language. Unscripted slice-of-life feeling. Strong environmental authenticity. Rich real-world details and believable human motion. Camera Style: Early-2000s consumer DV camcorder aesthetic. Friend casually recording everyday moments. Heavy handheld shake, imperfect framing, frequent autofocus hunting, lens breathing, exposure pumping when moving between sun and shade, occasional motion blur, subtle rolling shutter, mild digital compression artifacts, faded colors, soft contrast, slight sensor noise. No stabilization. No cinematic camera moves. No modern color grading. TIMED SCENES (00:00–00:15) 00:00–00:02 Outside a whitewashed village house. She sits on worn stone steps adjusting her loosely tied headscarf with both hands raised. Loose strands of dark hair move in a light breeze. She smiles naturally while the camera struggles to hold focus. 00:02–00:04 The camera follows her as she walks along a cobblestone path carrying an empty copper water jug, her skirt swaying with each step. Framing drifts off-center as the operator tries to keep up. 00:04–00:06 At the old stone village fountain, water flowing steadily from a carved stone spout into the basin. She rests one foot on the low stone edge and leans forward, holding the copper jug directly under the stream of water. The movement lifts the hem of her long skirt, and it slides up her raised leg, leaving her bare calf and knee visible for the rest of the shot. She doesn't notice. Natural morning light, no change in lighting. The autofocus drifts once toward her leg, then returns to the flowing water. The water keeps flowing as she lifts the jug away. 00:06–00:08 She lifts the full jug with visible effort and carries it back along the path, adjusting her grip. Chickens scatter near her feet. Exposure pumps as she passes between sunlight and the shade of a grapevine pergola. 00:08–00:10 Under the grapevine pergola beside her house. She reaches up on her toes to pick a small bunch of ripe grapes hanging overhead, dappled sunlight moving across her face. She lowers the bunch, plucks a single grape, and puts it in her mouth, smiling to herself. Loose handheld side angle with natural camera drift, autofocus hunting between the grape leaves and her face. 00:10–00:12 Close side profile as she stands under the pergola holding the grapes. She suddenly notices the camera has been recording, turns directly toward the lens with a surprised laugh, raises her hand, and warmly says, "Hoş geldin." The camera catches the moment slightly late. 00:12–00:15 She walks slowly down a poplar-lined dirt lane carrying the copper jug in one hand and the grapes in the other. She notices the camera again, gives a small shy smile, tucks a loose strand of hair under her scarf, then looks away and continues walking. Recording cuts abruptly to black mid-motion as if the camcorder was switched off. Audio: Natural ambient sound only — morning birds, a distant rooster, flowing fountain water, light wind through poplar leaves, chicken sounds, footsteps on cobblestone and dirt, fabric moving, a distant tractor, subtle village ambience. No music. No sound design. No narration. Goal: Authentic Anatolian village life captured like a forgotten home video from the early 2000s — candid, imperfect, realistic, warm, and deeply believable. Everyday moments of a young village woman doing ordinary tasks.

Cinematic scene

Main Subject: Young Korean woman, early 20s, sporty hiking outfit, cap over a ponytail, small backpack, water bottle in hand, refreshed cheerful expression. Location: Forest trail leading to a small waterfall, mid-morning. Mossy rocks, mist rising from the falling water, tall trees, dappled sunlight. No other hikers visible. Visual Style: Ultra-realistic documentary realism, refreshing outdoor feeling, bright natural forest light. Camera Style: Early 2010s digital camcorder, handheld with natural shake, autofocus hunting in shifting light and mist, soft grain, occasional lens flare. No stabilization. Timeline (15 sec, each slot = 2 compressed beats): 00:00–00:03 → She reaches the waterfall, then stops to take in the view, breathing heavily. 00:03–00:06 → She dips her hand into the cool spray, then says to camera "이거 진짜 시원하다" ("This feels really refreshing"). 00:06–00:09 → She sits on a mossy rock, sipping water from her bottle. 00:09–00:12 → She watches the mist drift, wiping sweat from her forehead. 00:12–00:15 → She looks at camera saying "여기까지 온 보람 있다" ("It was worth hiking all the way here"), smiling as it fades. Audio: Waterfall rushing, birds chirping, faint wind, water dripping on rocks. Her dialogue as noted above. No music. Goal: A refreshing, rewarding hiking rest moment energetic, warm, genuine.

Cinematic scene

Cinematic summer beverage commercial for FIZZI Sparkling Tonic — Grapefruit, high-end style in a warm golden-hour color grade with teal-turquoise pool contrast and coral-pink accents, mixing lifestyle wide shots with hyper-detailed macro product close-ups, slow-motion moments, fun youthful carefree energy, upbeat tropical pop and sound design of can hiss, carbonation fizz, pool splashes and laughter. The product is a tall slim coral-pink aluminum can with condensation droplets — bold retro-rounded lowercase magenta "fizzi" at the top, "sparkling tonic" below, a 4x4 pop-art grid of illustrated grapefruit circles in pink, coral, hot pink and pale yellow, with "0% Alcohol" and "GRAPEFRUIT" at the bottom. At golden hour on a stylish rooftop pool, a young woman in a coral-pink swimsuit sits at the pool edge with her feet in turquoise water, lifting an ice-cold FIZZI can to her lips and closing her eyes in pleasure, friends mingling and string lights twinkling in a dreamy soft-focus blur behind her, condensation catching the golden light. A slightly wider shot shows her floating on a clear inflatable beside a friend in a pale yellow swimsuit, both laughing as they clink their coral-pink cans with a small splash, shot at water level with sparkling reflections dancing across the frame. An extreme macro slowly pans across the coral-pink can surface in hyper-detail — beads of condensation catching prismatic light, the glistening magenta "fizzi" logo filling the frame, then down to the crisp pop-art grapefruit grid with shimmering turquoise bokeh behind. Another dreamy slow-motion macro shows tiny carbonation bubbles racing up through pale-pink sparkling liquid in a glass with a glowing grapefruit slice floating, the can softly out of focus behind, with ASMR fizzing crackle. Finally the camera pulls back to a wide golden-hour shot of the full party — both women floating with cans, friends dancing and lounging on coral towels, grabbing FIZZI from an ice bucket of grapefruit halves — then slowly pushes in as they raise their cans toward camera and clean white text fades in: "Grab an ice cold fizzi," with "© 2026 FIZZI Beverages Co." below, holding then fading out.

一句話瞭解

這個模型表現如何?

上傳一段來源影片和一張替換角色圖片,選擇解析度後即可建立角色替換任務。結果中的身分一致性、動作銜接和場景融合仍需逐段檢查。

它擅長什麼
如何使用 Wan 動畫替換? — 使用 Wan 2.2 Animate 在 Kyeo AI 以一段上傳影片和一張圖片替換主體;480p/580p/720p 按每個已驗證影片秒計費 6/9.5/12.5 credits。
適合哪些使用者
適合使用一組權利清楚的來源影片與角色圖片進行受控替換測試,並比較三檔解析度的實際結果;不適合用提示詞生成鏡頭、單圖動畫、影片續寫或時間軸編輯。
為什麼要在 Kyeo AI 使用
Kyeo 會顯示三檔每秒費率和固定輸入;送出前核對上傳素材、已驗證影片時長、檔案格式與解析度,缺少時長證明時會在扣點前拒絕。

關鍵參數總覽

協助你快速評估此模型是否符合你的業務情境。

模型類別
AI 影片模型
廠商
Alibaba
模型家族
Wan2.2-Animate
串接方式
目前站內串接
任務契約
Wan — 非同步任務
執行方式
非同步任務
輸入素材
來源影片、替換圖片與輸出解析度
素材邊界
影片 MP4/QuickTime/Matroska;圖片 JPEG/PNG/WebP;各 10 MB

其他名稱

同一個模型在官方文件、社群討論與 API 串接層可能有不同稱呼,這裡幫你統一整理。

Wan — Wan 動畫替換如何驗證影片時長?

使用者常見問題

從實際任務、輸入條件和結果要求出發,整理選擇模型時最需要確認的問題。

如何使用 Wan 動畫替換?
Wan 動畫替換需要上傳哪些素材?
Wan 動畫替換接受哪些影片與圖片格式?
Wan 動畫替換如何按影片秒計費?
Wan 動畫替換是否需要提示詞?
Wan 動畫替換會裁切替換圖片嗎?
Wan 動畫替換是否會檢查生成內容的安全性?
Wan 動畫替換如何驗證影片時長?
Wan 動畫替換與 Wan 2.2 A14B Turbo 有何不同?
Wan 動畫替換與 Wan 2.6 應如何選擇?

選擇指南

還在猶豫該用哪個模型嗎?先查看這幾個關鍵決策點。

1
先上傳兩類素材

必須透過工作台上傳一段來源影片和一張替換圖片;請求會驗證上傳歸屬,來源影片還必須具備可複核的時長證明。

2
再依費率選擇解析度

480p、580p、720p 分別按每個已驗證影片秒 6、9.5、12.5 credits 計費,Kyeo 會將請求總價向上取整。解析度不保證替換穩定、人物一致或邊緣品質。

3
送出後逐項人工複核

檢查替換對象、動作、遮擋、邊緣、背景、文字與音訊,並核對圖片縮放或置中裁切後的構圖;未以相同素材完成真實任務前,不得寫成效果保證。

使用者常搜尋的比較

把常見候選放在同一任務下比較,協助你確認輸入方式、控制項和成本差異。

Wan 動畫替換與 Wan 2.2 A14B Turbo 的輸入有何不同?
Wan 動畫替換 vs Wan 2.2 A14B Turbo:點數與參數
Wan 動畫替換與 Wan 2.6 的影片任務有何不同?
Wan 動畫替換 vs Wan 2.6:素材與解析度
Wan 動畫替換與 Wan 2.5 Video 的控制項有何不同?
Wan 動畫替換 vs Wan 2.5 Video:固定替換或生成

模型比較表

還在猶豫該選哪一個嗎?這張表協助你快速看清目前模型與替代方案的核心差異。

任務範圍
Wan 動畫替換
固定角色替換:單一影片 + 單一圖片
Wan 2.2 A14B Turbo
文字、單圖或圖片加音訊三路生成
Wan 2.6
文字、單圖或單一影片三路生成
輸入契約
Wan 動畫替換
上傳單一影片 + 單一圖片;無提示詞控制項
Wan 2.2 A14B Turbo
提示詞必填;依工作流程使用圖片或音訊
Wan 2.5 Video
提示詞必填;可選單圖
目前點數
Wan 動畫替換
6/9.5/12.5 × 已驗證影片秒
Wan 2.2 A14B Turbo
標準 40/60/80;語音 12/18/24 每秒
Wan 2.6
70-315 點數,依解析度與時長
可見控制項
Wan 動畫替換
一個影片上傳、一個圖片上傳與解析度
Wan 2.2 A14B Turbo
12 項,依文字、圖片、語音工作流程顯示
Wan 2.6
解析度與時長
提交驗證與安全設定
Wan 動畫替換
提交前驗證素材歸屬與影片時長;目前流程不提供提示詞或安全檢查開關
Wan 2.2 A14B Turbo
驗證站內上傳素材;語音工作流程預設啟用安全檢查
Wan 2.6
驗證站內上傳素材;目前不提供多鏡頭或安全檢查開關

實際使用體驗

結合公開資料、目前站內邊界和典型任務,說明模型適合與不適合的情況。

同一家族不等於同一服務契約

Wan 官方資料證明 Animate-14B 可進行 animation 與 replacement;Alibaba Cloud 直連另提供 `wan2.2-animate-mix`。Kyeo 使用目前公開資料介面工作流程,不能沿用直連的模式、長度、解析度或素材上限。

產品表單比行銷效果詞更接近目前邊界

頁面應以目前單一影片加單一圖片的上傳契約、影片時長門檻與目前公開資料欄位為準,不能依模型名稱推斷角色保留或動作品質。

功能詳解

固定的角色替換輸入

透過工作台上傳來源影片和替換圖片並選擇解析度;只有上傳歸屬與影片時長證明都通過驗證才會建立任務。

圖片比例處理

目前公開資料介面產品表單說明,替換圖片與目標比例不同時會縮放並置中裁切。Kyeo 沒有獨立比例控制,使用者應保留安全構圖並檢查實際裁切。

安全欄位邊界

安全檢查會隨任務啟用。生成結果仍需人工核對內容安全、素材授權、人物身分與發布合規。

適合做什麼

單人角色替換測試

選擇主體清楚、遮擋較少的短片與一張權利清楚的角色圖片,觀察替換範圍、動作連續性及背景是否受到影響。

素材構圖相容性比較

使用不同取景與人物比例的圖片測試縮放和置中裁切,確認頭部、身體與服裝是否仍位於安全區域。

三檔解析度同題複核

保持同一影片與圖片不變,只切換 480p、580p、720p,分別記錄輸出是否成功、耗時、邊緣與主體問題。

提示詞技巧

這條工作流程沒有提示詞

目前 Kyeo 頁面沒有提示詞輸入,目前公開資料介面現行工作流程表單也未把 prompt 列為預期欄位。

讓來源影片主體容易辨認

優先使用主體完整、切鏡較少且遮擋有限的影片。這項建議用於降低評審變因,不代表目前公開資料介面一定能穩定保留動作或背景。

準備可安全裁切的角色圖

讓關鍵人物內容靠近畫面中央並保留邊緣空間;目前公開資料介面可能縮放並置中裁切比例不符的圖片,送出前也應確認肖像與素材使用權。

選擇它的理由

固定的單一影片加單一圖片契約容易與通用生成模型區分。
三檔解析度的每秒費率與已驗證影片時長可在送出前直接核對。`duration ≤ 30s`。
目前公開資料介面的表單明確列出素材格式、大小與圖片裁切行為。
Wan 官方資料可以證明角色替換家族定位,不必借用無關影片能力。

使用前須知

來源影片和替換圖片必須由目前工作台上傳,使用者不能直接填寫任意遠端 URL;上傳歸屬或影片時長證明缺失時會在扣點前拒絕。
目前公開資料介面仍會在建立任務時確認素材的實際格式、大小與可存取性。
生成結果仍需檢查內容安全、素材授權、人物身分與發布合規。
角色替換、動作保留、裁切、遮擋、音訊、回呼與連結期限會受實際輸入和目前公開資料狀態影響;三檔費用以送出前顯示與任務紀錄為準。

可調整參數

協助你快速評估此模型是否符合你的業務情境。

來源影片
必填
參數類型: text
替換圖片
必填
參數類型: text
輸出解析度
選填
參數類型: select
預設值:480p
480p
580p
720p

點數消耗

6/9.5/12.5 點數/影片秒

目前公開資料服務依已驗證來源影片時長計費:480p 每秒 6 credits、580p 每秒 9.5、720p 每秒 12.5。Kyeo 對請求總價執行 Math.ceil;影片時長證明缺失或不符時會在扣點前拒絕。

省點數小技巧

批次生成圖片或影片前,建議先用同一組素材,在目前模型與同類候選模型之間進行 A/B 測試;確認效果後再批次生成,避免浪費點數。

常見問題

同類模型推薦

還沒決定該用哪個嗎?順便比較一下這幾個同類候選模型。

Wan 2.2 A14B Turbo

Wan 2.2 A14B Turbo 在 Kyeo 並非單一工作流程:沒有素材時使用文字生成,上傳一張圖片後改走圖片生成,同時提供一張圖片與一段音訊才會進入語音驅動。三條工作流程的解析度、點數與可見參數不同,選擇前應先確認素材組合。

AI 影片模型
標準 40/80;語音 12/18/24 每秒

Wan 2.6

不附媒體時使用文字生成影片,恰好 1 張圖片時使用圖片生成影片,1–3 個影片時使用影片編輯;圖片與影片不可混用。 預設為 1080p、5 秒、鏡頭結構關閉、內容過濾開啟。

AI 影片模型
70–315 點數/次

Wan 2.5 Video

瞭解 Wan 2.5 Video 在 Kyeo AI 的文字與單圖影片流程、5/10 秒、720p/1080p、提示詞差異及 60–200 點數計費。 Kyeo 依解析度與長度精確計費:720p 5 秒 60 點數、720p 10 秒 120 點數、1080p 5 秒 100 點數、1080p 10 秒 200 點數;預設 720p 5 秒為 60 點數。Wan 2.5 的 60/100/120/200 點數如何計算? — 瞭解 Wan 2.5 Video 在 Kyeo AI 的文字與單圖影片流程、5/10 秒、720p/1080p、提示詞差異及 60–200 點數計費。

AI 影片模型
每次 60/100/120/200 點數

資訊來源

頁面內容依據廠商資料、模型系列說明與 Kyeo AI 平台串接設定整理。

來源說明

使用 Wan 2.2 Animate 在 Kyeo AI 以一段上傳影片和一張圖片替換主體;480p/580p/720p 按每個已驗證影片秒計費 6/9.5/12.5 credits。 目前公開資料服務依已驗證來源影片時長計費:480p 每秒 6 credits、580p 每秒 9.5、720p 每秒 12.5。Kyeo 對請求總價執行 Math.ceil;影片時長證明缺失或不符時會在扣點前拒絕。

最後更新: 2026-08-27
Wan2.2 官方程式庫:Animate-14B
官方
查看原始來源
Alibaba Cloud:wan2.2-animate-mix 直連 API
官方
查看原始來源