nai-diffusion-4-5-curated

由 NovelAI 提供图像

NAI Diffusion 4.5 Curated 可依据正向提示词、负向提示词和参考图生成插画、漫画和角色。

更新于:
输入提示词文本
模型4款可用
返回方式异步结果链接
价格8 quota / 张
Smiling girl in a yukata making a peace sign beneath fireworksFireworks yukata示例 01
Close-up portrait of a silver-haired girl in a librarySilver-haired scholar示例 02
Freckled blue-haired cat girl sitting in a softly painted barPastel cat girl示例 03
Chibi octopus in sunglasses and a sun hat riding a tidal waveTidal-wave octopus示例 04
生成示例

从提示词到成图

1girl, yukata, peace sign, fireworks, night, sparkles, watercolor, painterly, smile

solo, no humans, octopus, sunglasses, chibi, disgustingly adorable, lifebuoy, sun hat, tidal wave, traditional media, calligraphy brush (medium), no lineart, flat color, motion blur, from side, location

2girls, text, english text, location, -1::monochrome, speech bubble, simple background ::, looking at another, masterpiece, best quality, rating:general. The blonde girl is excitedly asking "How are you?" and the purple haired girl is responding "Fine, thanks" with a bored expression on her face. Text: How are you? Fine, thanks | 1girl, purple eyes, short hair, ruffled blouse, red blouse, blonde hair, green scarf, blunt bangs, fang, small breasts, long sleeves, bob cut | 1girl, very long hair, purple hair, curly hair, golden shirt, full body, green eyes, turtleneck sweater, sleeveless turtleneck, side braid, medium breasts, sleeveless

1girl, masterpiece, best quality, dynamic angle, close-up, long eyelashes, grey gradient hair, silver hair, ahoge, medium updo, thick eyebrows, hair between eyes, ponytail, pink blush, grey eyes, pink eyeshadow, smug, pink lips, glossy lips, short hair, white hair, messy hair, medium hair, ponytail, makeup, pink makeup, dark blue top, high ponytail, puffy hair, hair focus, very aesthetic, location, masterpiece, library background, location

1boy, high quality, {{masterpiece, very aesthetic}}, {traditional media}, dynamic pose, dynamic angle, glitter, handsome, male focus, broad shoulders, [smug, painterly], year 2020, broad shoulders, exquisite clothes, abstract background, royalty, beautiful young man, scepter, young, location

1girl, character focus, close-up to the girl, cover illustration, three quarter view, key visual, from below, star-shaped pupils, knee bent, jumping, dutch angle, cowboy shot, anime coloring, wearing flowing blonde hair, glowing aura covering whole body, rim lighting, sunlight, cinematic color grading, wearing multi-layered blue and white tone long dress, off shoulder, princess crown, dramatic cloudy blue sky, floating, -3::simple illustration::, -2::simple background::, -3::minimalism::, year 2024, 1.3::pale color::, overexposure, magnificent scenery, large title text on the bottom of the frame, Title says "NAI Diffusion V4.5Curated". Text: NAI Diffusion V4.5 Curated

1girl, cat girl, topknot, light blue hair, orange eyes, skinny, freckles, black skirt, tan, red tank top, bar, bored, feet out of frame, traditional media, pastel (medium), depth of field, location

1girl, best quality, firefly princess, castle background, dark, wing crown, dress flower, glowing, butterfly hair ornament, flowers, petals, artbook, muted colors, watercolor pencil (medium), red light, looking away, nose, location

1girl, very aesthetic, best quality, amazing quality, colorful, cowboy shot, magic, swirls, sword, scabbard, holding book, pointing, open book, knight, long bangs, armor, gauntlets, wizard, witch hat, hat feather, feather, robe, quilted vest, leather belt, leather pouch, serious, medieval, forest, white hair, blue eyes, shaded face

模型能力

nai-diffusion-4-5-curated 能做什么

根据提示词生成

提交简洁提示词,通过一个 TTAPI 任务获取图像结果。

常用参数

设置提示词、模型、输出参数和回调即可接入。

异步查询

通过轮询或 Webhook 接收结果,避免耗时生成阻塞界面。

接入产品

NovelAI 的输出可以展示、保存或交给后续流程处理。

计费

按操作计费,单位为 quota

按模型、操作和输出设置扣除 quota。

图像生成官方定价
模型价格计费单位
NAI Diffusion 5 Full · Curated8 quota / 张
NAI Diffusion 4.5 Full · Curated8 quota / 张
API 请求

请求头与提交参数

使用 TTAPI 密钥,按下方字段向接口发送请求。

POSThttps://api.ttapi.io/novelai/text-to-image

请求头

TT-API-KEYstring必填

填写 TTAPI API 密钥。

Content-Typestring必填

使用 application/json。

请求体

promptstring必填

描述想在插画中出现的内容。

modelenum必填

可选 nai-diffusion-5-full、nai-diffusion-5-curated、nai-diffusion-4-5-full、nai-diffusion-4-5-curated,默认 nai-diffusion-5-curated。

ninteger选填

生成图像数量,范围 1 至 4,默认 1。

widthinteger选填

输出宽度,单位为像素,默认 1024。

heightinteger选填

输出高度,单位为像素,默认 1024。

seedstring选填

复现或改变生成结果时使用的随机种子。

stepsinteger选填

扩散步数,默认 23。

scalenumber选填

提示词引导强度,默认 5。

negative_promptstring选填

描述要避免的元素或效果。

quality_toggleboolean选填

启用文档所述的质量预设,默认 true。

samplerenum选填

可选 Euler Ancestral、Euler、DPM++ 2S Ancestral、DPM++ 2M SDE、DPM++ 2M、DPM++ SDE,默认 Euler Ancestral。

noise_scheduleenum选填

可选 karras、exponential 或 polyexponential,默认 karras。

character_promptsobject[]选填

角色提示词与位置设置。

vibe_transferobject[]选填

氛围参考图及影响强度。

character_referenceobject[]选填

角色和风格参考设置。

hook_urlstring选填

任务完成时调用的回调 URL。

查看 TTAPI 接入文档
接入示例

发送请求并获取结果

复制当前模型的请求地址、鉴权请求头和请求体。

curl --request POST \
  --url https://api.ttapi.io/novelai/text-to-image \
  --header "TT-API-KEY: $TTAPI_KEY" \
  --header 'Content-Type: application/json' \
  --data '{"prompt":"anime character portrait, detailed eyes","model":"nai-diffusion-4-5-curated","n":1,"width":1024,"height":1024,"steps":23,"scale":5,"quality_toggle":true,"sampler":"Euler Ancestral","noise_schedule":"karras","hook_url":"https://example.com/webhooks/novelai"}'