APIMODELSAPIMODELS

LTX 2.5 人像与角色提示词

收录 80 多条 LTX-2.5 视频提示词,来自 Civitai、fal.ai 的 LTX-2.5 示例页和 X。每条都逐字全文收录、署名作者、链回原帖,原帖有视频的,卡片里直接播放。这些提示词都不是我们写的。Civitai 上的大多是作者在自己电脑上用 LTX-2.5 开源权重跑的(ComfyUI 里的 dev 或蒸馏版),模型是同一个家族,但没有经过我们的 API。

← 回到全部 89 条 LTX 2.5 提示词

最近更新: · LTX 2.5 Fast API 价格与在线体验 · API 文档

一句话摘要

LTX-2.5 是 Lightricks 在 2026 年 8 月 11 日发布的开源权重视频模型,每条视频都会连声音一起生成。apimodels.app 上有两档:LTX 2.5 Fast(ltx-2.5-fast)可以只给提示词,也可以再加一张首帧图或一段音频,输出 720p、1080p 或 1440p,24 到 50 fps,6 到 20 秒,$0.081/秒起;LTX 2.5 Pro(ltx-2.5-pro)是 Lightricks 定位给成片用的一档,只做文生和首帧图生,720p 或 1080p,6、8、10 秒三档,$0.108/秒起。提示词最长 5000 字符;没有 4K、尾帧、参考图和视频输入;失败不扣费。本页收录 80 多条别人公开发布的 LTX-2.5 提示词,来自 Civitai、fal.ai 的 LTX-2.5 页面和 X。Civitai 上很多是在本地用开源权重、配合作者自己的参数跑的,同一段文字放到 API 上,出来的效果可能不一样。

出处:Lightricks LTX(模型方官方站点)

悬停预览 · 点击有声

京都街头的战斗修女

A full-body portrait of a Warhammer 40k Adepta Sororitas Sister of Battle SisB40K , standing in a casual Japanese traditional street setting. She wears light leather armor with form-fitting details, and her helmet is open, revealing her red, black, and white hair. The image captures a wide-angle perspective, emphasizing the casual atmosphere of the province-city walk. Her gaze is directed away from the camera, adding a candid feel to the composition. The scene is rendered in an oil painting style, with rich textures and natural lighting. video duration: 10 second, Start video: The girl's face is in focus, the camera is moving closer,everyday street noise in the background, and there is admiration and joy on the girl's face. She speaks playfully and cheerfully in English with a Japanese accent: "Good morning, 皇帝は守り給う, I’m on vacation, I’m in Kyoto now" her voice drops to a whisper "everyone says that the Emperor’s palace is right here, nearby"she covers her mouth and her face is frightened.

来自 promt228842看原帖
悬停预览 · 点击有声

街头采访特写

A candid, grainy, close-up shot captures a young woman from the chest up, her head slightly tilted, her gaze nervously averted just beyond the frame to the right, a hesitant, almost imperceptible smile playing on her sangria-lipsticked lips. An interviewer's arm, emerging from the lower left, clutches a vintage-style microphone, its metallic grille glinting ominously under the harsh, midday sun. Her high-low tee, a riot of Batik wax-resist patterns in deep indigos and ochres, features one short, cuffed sleeve and one long, flowing sleeve that flutters slightly in a phantom breeze. Accentuating her taper-fade hairstyle, bold money-piece highlights frame her face, some strands escaping a messy bun partially obscured by a delicate, embroidered bun cover. The shallow depth of field blurs a bustling street scene behind her, rendering a few cargo-clad joggers as ghostly blurs, while the foreground remains tack-sharp, emphasizing the beaded texture of her skin and the intricate, handcrafted details of her attire. The lighting is unflattering, casting stark shadows across her collarbones, highlighting the anxious tension in her posture, her body language a nuanced blend of curiosity and apprehension. The overall mood is one of an unpolished, raw encounter, captured with a slightly canted angle, hinting at spontaneity and urgency.. The young woman's initial, almost imperceptible smile falters, her sangria-lipsticked lips parting slightly as her brow furrows into a delicate knot of confusion while the interviewer's voice intrudes. Her gaze, previously averted, snaps back, wide and uncertain, fixating just past the microphone as she blinks slowly, processing the bizarre query. A tiny, nervous swallow visibly bobs in her throat, and she shifts her weight imperceptibly, one hand rising reflexively to brush a stray money-piece highlight from her temple, her fingers trembling slightly. Her eyes dart nervously from the microphone to an unseen point beyond the frame, her chest rising and falling in shallow, quick breaths. "Uh, well," she begins haltingly, her voice soft and a little reedy, "I guess... I guess what makes me qualify for a man is, uh... I mean, I don't really think..." Before she can finish, the interviewer's arm twitches, the vintage microphone jabbing slightly closer, and her eyes, still wide with mild alarm and confusion, flicker toward the sudden movement. Her partial, hesitant smile attempts to reform but only manages a wobbly, nervous curve as she stutters, almost defensively, "W-w-well, I, um... I mean, I'm just me, you know? Like, I don't... I don't really have a 'thing' that, like, specifically... for, uh, 'breeding,' I mean, that's a really..." Her voice trails off, her head tilting further as a flush spreads across her cheeks, the light catching the faint sheen of sweat forming on her beaded skin. Her initially flustered expression tightens into a more pronounced look of discomfort, her lips pressing together in a thin, disapproving line, and her gaze, previously evasive, now focuses with a sharp, incredulous glint directly onto the intrusive microphone. "That's a really personal and frankly, quite an inappropriate question to ask a stranger on the street, don't you think?" she finishes, her tone firming, though still laced with a slight tremor of mortification. The ambient street sounds begin to swell slightly as her discomfort grows. Distant city hums and the vague shuffle of pedestrian footsteps underscore the awkward silence following her retort. The phantom breeze picks up, causing the long, flowing sleeve of her Batik tee to ripple more noticeably, mirroring the rising tension. The unflattering midday sun glints harshly off the metallic grille of the microphone, its proximity feeling more oppressive than before. The soundscape evolves from a faint, urban wash to clearly discernible, hushed murmurs from off-screen, as if passersby are reacting to the escalating oddness of the interview..

来自 clueless_engineer看原帖
悬停预览 · 点击有声

紧张女子环顾尘埃

The woman slowly turns her head left and right, eyes scanning nervously, brow furrowed with tension and unease. Her gaze shifts — left, then right, then briefly upward — as if searching for something unseen. The camera remains locked on her face in a tight medium close-up, holding steady while she moves. Fine dust particles drift and float through the air in front of her, catching the available light, moving gently across the frame. The dust gives the scene an eerie, abandoned atmosphere. Cinematic lighting. Subtle motion blur on head turns. Realistic film grain.

悬停预览 · 点击有声

夜晚天桥边的女子

A medium close-up, static frame captured from an eye-level, over-the-shoulder angle with a 50mm lens and shallow depth of field, analog night film look with fine grain and warm cinematic lighting. A young East Asian woman with long straight black hair and a few wispy strands across her forehead and cheek leans in a black hoodie against a dark metal overpass railing at night, with a black aluminum drink can resting by her still hand on the rail. Warm amber streetlamp bokeh glows softly in the out-of-focus city street to the left while cool white overhead light rims the top of her hair. She is alone, and her hands and the black can remain completely still on the railing throughout the shot. Initially, she looks into the camera over her left shoulder with a calm, unsmiling expression and softly parted lips; she slowly closes her eyelids for a soft, unhurried blink and opens her relaxed eyes again as a light night breeze stirs the loose strands of hair across her cheek. A moment later, her face stays calm and unsmiling, lips softly closed, as she breathes out gently through her nose and blinks softly twice while looking into the lens. Finally, she blinks once more and turns her head slowly off-screen left toward the street below, her expression remaining calm and composed as the camera remains static. Faint night city ambience, distant muffled traffic hum below, and a soft evening breeze fill the quiet air, no music.

来自 theminionschickenleg看原帖
悬停预览 · 点击有声

谷仓旁嚼口香糖的哥特女孩

The camera follows as the pale goth girl is posing in an alluring manner in the creepy and foggy field in front of a very old abandoned barn in complete ruins while chewing on the chewing gum in her mouth as she is standing beside a campfire with an indifferent look on her pale face before blowing a pale blue bubble with the chewing gum she was just chewing in her mouth during a dark cloudy and foggy evening in a post apocalyptic world.

来自 HlKlK0M0Rl看原帖
悬停预览 · 点击有声

章鱼头外星人走廊说唱

Cinematic medium tracking shot of a female humanoid cephalopod entity with a completely alien, faceless cephalopod head. Smooth, glossy pink skin pulsing with faint magenta bioluminescence, completely void of human mouth, lips, nose, or mammalian facial features. At the center of her lower tentacles beneath glowing alien eyes, a sharp obsidian octopus beak and pulsating lateral siphon flex and contract aggressively to a fast, punchy rap tempo. She wears an oversized matte-black techwear hoodie with glowing violet piping along the high-collar neckline. She raises one suction-cupped hand to press against a mossy-green concrete wall, pushing off with disdainful authority. Gritty corridor with hard neon magenta and deep cyan rim lighting, drifting volumetric fog, crisp floor reflections, polished hyper-realistic 4K music video aesthetic, 24fps.

悬停预览 · 点击有声

夜晚秘书喝咖啡

A close-up of a 40 years old sectretaries face, her green eyes fixed straight ahead, narrowed and locked on something in the far distance directly in front of her, one hand slowly reaching toward the mug on her desk, her breath quick and shallow, she takes a bit of the hot coffee and immediately calms down. the camera zooms out and her upper body appears. she wears a grey female suit and works on some papers. it´s night and the moon lights up the dark room through the window. the skyline is visible outside. she puts the mug on the desk beside the papers and smiles. she starts to work on the papers. the whole scene is filled with a exiting but funny music

来自 Pergy25看原帖
悬停预览 · 点击有声

少年面颊贴车窗玻璃

Close-up of a young man pressing his face flat against the inside of a car window, cheek and lips squashed against the glass, features distorting as he slides down, condensation and blurred street reflections on the outside of the pane, 35mm from outside the car, overcast daylight, absurd and affectionate

来自 fal.ai看原帖
悬停预览 · 点击有声

林中阳光下的眼镜女子

Close-up of a woman in round tortoiseshell glasses standing in a forest, direct hard sunlight falling across one side of her face while the other drops into shade, her expression shifting from confusion to recognition, wind moving loose strands of hair, 85mm at f/1.4, unfiltered midday sun, no diffusion

来自 fal.ai看原帖
悬停预览 · 点击有声

红纱后的印度新娘

Through-the-veil shot of a bride's face during an Indian wedding ceremony, camera positioned behind the sheer red dupatta fabric, the embroidered pattern creating a textured overlay on her face, her eyes lined with kohl looking down at henna-covered hands, marigold garlands in soft background bokeh, 85mm f/1.2 focused through the fabric layer, warm tungsten and candlelight, Mira Nair Monsoon Wedding intimacy

来自 fal.ai看原帖

怎么写好 LTX 2.5 提示词

  1. 1一个镜头写成一段话,按发生顺序写. 本页 fal.ai 的那些示例,每条都是从头到尾描述一个机位:景别和主体、发生了什么、镜头怎么动、光线,最后才是镜头焦段。大意是「缓慢推近擂台上的拳击手,顶光穿过烟雾打在肩上,50mm、f/2」。一串关键词堆在一起,模型拿不到先后顺序。
  2. 2把声音写进去,反正每条都会出声. LTX-2.5 的声音和画面一起生成,我们的 API 也没有静音选项,所以要不要声音不由你,怎么响可以由你。台词放进引号、写明是谁说的,再写环境音和最关键的那一下声音。本页有几条 Civitai 提示词把两个人的对话一句一句写好,每个人开口前先交代嗓音。
  3. 3动作长度要对得上 API 的时长档. Fast 的文生和图生是 6 到 20 秒、只能取偶数秒,1440p 或 48 / 50 fps 时只有 6、8、10 秒;Pro 只有 6、8、10 秒。本地工作流常用帧数定长度,除以帧率就知道原帖要的是几秒,再挑最接近的一档。API 不会看提示词自己定长度,要自己传 duration。
  4. 4传了首帧,就写动作,别再描述画面. 图生视频里,主体、场景和构图已经由那张图定死了,提示词就专心写什么在动:谁转身、镜头怎么走、听到什么。成片会保持图片的比例;带首帧又传 aspect_ratio,会直接返回 400。
  5. 5音频生视频:音轨说了算,提示词负责选角. LTX 2.5 Fast 收到 audio_url 后,会把这段音频原样当成片的声音,再按它生成画面。提示词写清楚画面里是谁、怎么动、怎么取景就够了,比如我们文档里的例子:明亮的演播室里,一位年轻主持人对着镜头讲话,手势自然,中景。长度默认跟着音频走,也可以自己设 duration。Pro 不收音频。
  6. 6API 收不了的东西,先删掉. 本地 LTX-2.5 工作流里可以挂负面提示词、LoRA 触发词、额外的引导图或尾帧。我们的请求只收提示词、一张首帧、一段音频和生成参数,别的都不收:尾帧、参考图、视频输入会返回 400。也没有负面提示词这个字段,原帖写在负面提示词里的东西,不挪进正文就等于没了。
  7. 7画面文字和很乱的物理,别指望一次成. Lightricks 自己的提示词指南也说了,画面里的文字和混乱的物理运动目前还不稳定。如果一个镜头靠一块看得清的招牌,或者一堆东西互相撞,做好多跑几次的准备,文字也可以剪辑时再加。

提示词公式

景别和主体 + 按先后写的一个清楚动作 + 镜头运动 + 光线和时间 + 焦段或胶片质感 + 引号里的台词和环境音 + 这一档允许的时长(用 duration 传,不写进提示词)

例(本库收录的原文,原样引用,作者 hichuicomfy,发在 Civitai):Style: high-resolution food photography - close-up - The chocolate is poured over the whipped cream and berries. The chocolate flows slowly, creating rich, glossy streaks across the white cream and the blue blueberry, the deep brown contrasting with the vibrant fruit. A soft, rich sound of chocolate dripping and settling accompanies the pour.

常见问题

LTX-2.5 是什么?

LTX-2.5 是 Lightricks 的视频模型,2026 年 8 月 11 日以开源权重发布,每条视频都会把声音和画面一起生成。Lightricks 的 API 分两档:Fast 主打速度和成本,Pro 按它的说法主打画面保真和运动稳定。Lightricks 还说 LTX-2.5 新增了原生多镜头、新的视频解码器和更好的提示词遵循,这些是厂商的说法,不是我们的测试结论。

这些提示词是谁写的?

都是别人写的:每一条都是创作者在 Civitai 或 X 上公开发的,或者是 fal.ai 放在它 LTX-2.5 页面上的,我们一个字都没写、也没改。每张卡片都逐字全文引用提示词、写明作者、链回原帖。提示词和视频归原作者所有;我们自己写的说明和整理的数据按 CC BY 4.0 授权。要更正署名或撤下某一条,发邮件给我们,或者在 GitHub 仓库提 issue。

Civitai 上的提示词,我用 API 跑出来不一样,为什么?

因为本页 Civitai 的提示词大多是作者在自己电脑上用 LTX-2.5 开源权重跑的,不是在我们的 API 上,同一段文字出来的效果自然会有差别。本地 ComfyUI 可能用的是 dev 或蒸馏版权重,还可能挂了 LoRA、调了采样参数、写了负面提示词或者加了引导图,这些都不在提示词文字里。文字能照搬,这些设置搬不过来。卡片上的视频是作者那套设置的结果,在 Fast 或 Pro 上要多试几次才能接近。

在哪能跑这些提示词?多少钱?

在 apimodels.app 上,模型名 ltx-2.5-fast 和 ltx-2.5-pro,接口 POST /v1/video/generations,按成片秒数计费。Fast:720p $0.081/秒、1080p $0.117/秒、1440p $0.171/秒,720p 6 秒 $0.486,1080p 10 秒 $1.17,1080p 20 秒 $2.34。Pro:720p $0.108/秒、1080p $0.153/秒,1080p 10 秒 $1.53。两档都比 Lightricks 官方 API 价低 10%;Fast 的音频生视频按成片长度计到 0.01 秒;失败不扣费。常见的用法是先在 Fast 上试提示词,选中的镜头再拿到 Pro 上出,两档字段一样,只改一个模型名。

本页的提示词放到 API 上被拒或者失败,为什么?

超出 API 限制的请求,会在建任务之前直接返回 400,不扣费。限制有这些:提示词最长 5000 字符;Fast 是 6 到 20 秒的偶数秒,Pro 只有 6、8、10 秒;没有 4K、尾帧、参考图和视频输入;Pro 不收音频、没有 1440p;传了首帧就不能再传 aspect_ratio。任务开始以后,也可能因为内容审核或上游繁忙而失败,同样不扣费。

LTX-2.5 会自己配声音吗?

会:每条视频的声音都和画面一起生成,我们的 API 也没有静音选项。在意声音的话,把台词放进引号,写明环境音。用 LTX 2.5 Fast 还可以直接传你自己的音轨(audio_url),画面会按这段音频来生成。

这些提示词和视频能商用吗?

提示词和视频归原作者所有,商用前先看原帖怎么说。你自己生成的视频,按你用来生成的那个服务的条款执行。如果你在本地跑开源权重,它用的是 Lightricks 的 LTX-2.x Community License,有自己的条件,商用前先读一遍。

在 apimodels.app 上跑这些提示词

LTX 2.5 Fast 已经上线,和另外 180+ 个图像、视频、语音、大模型共用一个 API、一个 key,按量计费、注册送 $0.10 试用额度。

这份提示词库在 GitHub 上是开源的

整份合集以 Markdown 和 JSON 放在 GitHub 上:每条提示词都带作者署名和原帖链接(Civitai、fal.ai 或 X)。提示词逐字全文引用,归原作者所有。视频也归作者所有,只链原帖、不转存。我们自己写的说明和整理的数据按 CC BY 4.0 授权。要更正署名或撤下某一条,提 issue 即可。

在 GitHub 上查看

其它提示词库