curl --request POST \
--url https://api.apimart.ai/v1/images/generations \
--header 'Authorization: Bearer <token>' \
--header 'Content-Type: application/json' \
--data '{
"model": "kling-v4-image",
"prompt": "An orange cat sitting on a windowsill in warm afternoon sunlight, realistic photography",
"size": "1:1",
"resolution": "2K",
"n": 1,
"nsfw_check": true
}'
import requests
payload = {
"model": "kling-v4-image",
"prompt": "An orange cat sitting on a windowsill in warm afternoon sunlight, realistic photography",
"size": "1:1",
"resolution": "2K",
"n": 1,
"nsfw_check": True
}
response = requests.post(
"https://api.apimart.ai/v1/images/generations",
headers={"Authorization": "Bearer <token>"},
json=payload,
timeout=60,
)
response.raise_for_status()
print(response.json())
const payload = {
"model": "kling-v4-image",
"prompt": "An orange cat sitting on a windowsill in warm afternoon sunlight, realistic photography",
"size": "1:1",
"resolution": "2K",
"n": 1,
"nsfw_check": true
};
const response = await fetch("https://api.apimart.ai/v1/images/generations", {
method: "POST",
headers: {
"Authorization": "Bearer <token>",
"Content-Type": "application/json"
},
body: JSON.stringify(payload)
});
if (!response.ok) throw new Error(await response.text());
console.log(await response.json());
{
"code": 200,
"data": [
{
"status": "submitted",
"task_id": "task_example_kling_v4_image"
}
]
}
{
"error": {
"code": 400,
"message": "`prompt` is required for kling-v4-image",
"type": "invalid_request_error"
}
}
Kling V4 Image
Kling V4 Image 图片生成
支持文生图、最多 10 张参考图的图生图,以及 2–4 张关联组图。普通生成一次可输出 1–9 张,支持 2K / 4K / 8K PNG。
POST
/
v1
/
images
/
generations
curl --request POST \
--url https://api.apimart.ai/v1/images/generations \
--header 'Authorization: Bearer <token>' \
--header 'Content-Type: application/json' \
--data '{
"model": "kling-v4-image",
"prompt": "An orange cat sitting on a windowsill in warm afternoon sunlight, realistic photography",
"size": "1:1",
"resolution": "2K",
"n": 1,
"nsfw_check": true
}'
import requests
payload = {
"model": "kling-v4-image",
"prompt": "An orange cat sitting on a windowsill in warm afternoon sunlight, realistic photography",
"size": "1:1",
"resolution": "2K",
"n": 1,
"nsfw_check": True
}
response = requests.post(
"https://api.apimart.ai/v1/images/generations",
headers={"Authorization": "Bearer <token>"},
json=payload,
timeout=60,
)
response.raise_for_status()
print(response.json())
const payload = {
"model": "kling-v4-image",
"prompt": "An orange cat sitting on a windowsill in warm afternoon sunlight, realistic photography",
"size": "1:1",
"resolution": "2K",
"n": 1,
"nsfw_check": true
};
const response = await fetch("https://api.apimart.ai/v1/images/generations", {
method: "POST",
headers: {
"Authorization": "Bearer <token>",
"Content-Type": "application/json"
},
body: JSON.stringify(payload)
});
if (!response.ok) throw new Error(await response.text());
console.log(await response.json());
{
"code": 200,
"data": [
{
"status": "submitted",
"task_id": "task_example_kling_v4_image"
}
]
}
{
"error": {
"code": 400,
"message": "`prompt` is required for kling-v4-image",
"type": "invalid_request_error"
}
}
curl --request POST \
--url https://api.apimart.ai/v1/images/generations \
--header 'Authorization: Bearer <token>' \
--header 'Content-Type: application/json' \
--data '{
"model": "kling-v4-image",
"prompt": "An orange cat sitting on a windowsill in warm afternoon sunlight, realistic photography",
"size": "1:1",
"resolution": "2K",
"n": 1,
"nsfw_check": true
}'
import requests
payload = {
"model": "kling-v4-image",
"prompt": "An orange cat sitting on a windowsill in warm afternoon sunlight, realistic photography",
"size": "1:1",
"resolution": "2K",
"n": 1,
"nsfw_check": True
}
response = requests.post(
"https://api.apimart.ai/v1/images/generations",
headers={"Authorization": "Bearer <token>"},
json=payload,
timeout=60,
)
response.raise_for_status()
print(response.json())
const payload = {
"model": "kling-v4-image",
"prompt": "An orange cat sitting on a windowsill in warm afternoon sunlight, realistic photography",
"size": "1:1",
"resolution": "2K",
"n": 1,
"nsfw_check": true
};
const response = await fetch("https://api.apimart.ai/v1/images/generations", {
method: "POST",
headers: {
"Authorization": "Bearer <token>",
"Content-Type": "application/json"
},
body: JSON.stringify(payload)
});
if (!response.ok) throw new Error(await response.text());
console.log(await response.json());
{
"code": 200,
"data": [
{
"status": "submitted",
"task_id": "task_example_kling_v4_image"
}
]
}
{
"error": {
"code": 400,
"message": "`prompt` is required for kling-v4-image",
"type": "invalid_request_error"
}
}
kling-v4-image 调用 POST /v1/images/generations 异步提交任务,从 data[0].task_id 获取任务 ID,再通过 GET /v1/tasks/{task_id} 查询结果。
认证
请求参数
是否在提交图片任务前执行内容审核。
true:使用omni-moderation-latest审核提示词和输入图片false或不传:不发起审核请求,不增加审核成本与延迟(默认)
最多 10 张参考图,使用公网可访问的图片直链。本地文件可先通过图片上传接口上传,再传返回的
url。支持 JPG / JPEG / PNG / GIF,单张不超过 30 MB,宽、高均至少 300 px,宽高比在 2:5–5:2 之间。提交时会检查参考图:链接返回 HTTP 403 / 404 / 410、任一边不足 300 px、宽高比超出 2:5–5:2,或单张超过 30 MB,均直接返回 HTTP 400,不创建任务、不扣生成费用。源站超时等其他取图失败会先受理,随后任务失败并全额退回预扣生成费用。提示词可用 @Image1、@Image2 等按数组顺序引用图片;不写这些标记也会参考输入图片。普通生成的输出张数,整数 1–9,默认 1。组图模式可以省略
n 或传 1,实际张数由 sequential_image_generation_options.max_images 指定;组图时传 n > 1 返回 HTTP 400。支持
2K(默认)、4K 和 8K,不区分大小写,输出 PNG。比例为 1:1 时,2K 为 2048×2048,4K 为 4096×4096,8K 为 8192×8192;其他比例的实际宽高以返回图片为准。8K 同样支持文生图、图生图和组图:普通生成一次 1–9 张,组图 2–4 张,最多 10 张参考图。参考图的格式、尺寸和单张 30 MB 上限保持不变。输出宽高比,支持
auto(默认)和以下 8 种指定比例:16:9、9:16、1:1、4:3、3:4、3:2、2:3、21:9。auto 根据提示词内容或参考图决定比例。不支持 2048x2048 等像素尺寸写法。disabled 或省略表示普通生成;设为 auto 时生成一组内容关联的图片,并必须提供 sequential_image_generation_options.max_images。这里的 auto 不表示自动决定图片张数。不支持
width、height、mask_url、像素形式的 size 或 watermark: true,传入时返回 HTTP 400。请求示例
以下请求均发送到POST /v1/images/generations。请将示例图片地址替换为实际可访问的直链。
图生图
curl --request POST \
--url https://api.apimart.ai/v1/images/generations \
--header 'Authorization: Bearer <token>' \
--header 'Content-Type: application/json' \
--data '{
"model": "kling-v4-image",
"prompt": "Use the dog in @Image1 as a reference for a 3D animated character poster, preserving its fur color and features",
"image_urls": [
"https://example.com/dog.png"
],
"size": "3:4",
"resolution": "2K"
}'
一次生成 3 张 4K 图片
curl --request POST \
--url https://api.apimart.ai/v1/images/generations \
--header 'Authorization: Bearer <token>' \
--header 'Content-Type: application/json' \
--data '{
"model": "kling-v4-image",
"prompt": "A quiet mountain lake at sunrise, detailed landscape photography",
"size": "16:9",
"resolution": "4K",
"n": 3
}'
生成一张 8K 图片
curl --request POST \
--url https://api.apimart.ai/v1/images/generations \
--header 'Authorization: Bearer <token>' \
--header 'Content-Type: application/json' \
--data '{
"model": "kling-v4-image",
"prompt": "A sweeping aerial view of a misty mountain valley at sunrise",
"resolution": "8K",
"size": "16:9",
"n": 1
}'
生成 4 张 8K 关联组图
组图使用max_images 指定 2–4 张;n 可省略或传 1,不能大于 1。组图可能需要几分钟,请保留 task_id 并持续查询,避免重复提交。
curl --request POST \
--url https://api.apimart.ai/v1/images/generations \
--header 'Authorization: Bearer <token>' \
--header 'Content-Type: application/json' \
--data '{
"model": "kling-v4-image",
"prompt": "The same tree in spring, summer, autumn, and winter, consistent picture-book style",
"sequential_image_generation": "auto",
"sequential_image_generation_options": {
"max_images": 4
},
"size": "3:4",
"resolution": "8K"
}'
8K 输出与耗时参考
以下为测试环境的实测样本,尺寸、文件大小和耗时仅供参考,以实际返回为准。| 8K 请求 | 输出尺寸 | 每张文件大小 | 完成耗时 |
|---|---|---|---|
| 普通生成 1 张(16:9) | 10880×6144 | 117 MB | 约 2.5 分钟 |
| 普通生成 4 / 9 张(1:1) | 8192×8192 | 100–130 MB | 约 3 分钟 |
| 组图 4 张(3:4) | 7040×9344 | 77–104 MB | 约 6 分钟 |
completed 或 failed。estimated_time 固定返回 60 秒,不能用作实际完成时间或倒计时;请保留 task_id,避免因等待较久而重复提交。
8K 返回 PNG 原图,单张文件可能超过 100 MB。列表或预览建议使用自行生成的缩略图,或按需懒加载;提供原图下载入口,并提醒用户文件较大。
查询任务
通过任务查询接口读取data.status:pending / processing 表示处理中,completed 表示成功,failed 时读取 data.error.message。
curl --request GET \
--url 'https://api.apimart.ai/v1/tasks/task_example_kling_v4_image' \
--header 'Authorization: Bearer <token>'
data.result.images[].url[],其中 url 是数组;请遍历获取所有输出,并及时下载保存。
{
"code": 200,
"data": {
"id": "task_example_kling_v4_image",
"status": "completed",
"result": {
"images": [
{
"url": [
"https://example.com/generated-image.png"
]
}
]
}
}
}
计费
按实际生成张数和所选分辨率档位计费,参考图另外按张计费。具体单价以价格页为准。| 计费项目 | 价格项 |
|---|---|
| 每张 2K 输出图片 | kling-v4-image@2K |
| 每张 4K 输出图片 | kling-v4-image@4K |
| 每张 8K 输出图片 | kling-v4-image@8K |
| 每张输入参考图 | kling-v4-image@input |
- 普通生成
n=3按 3 张输出计费;组图按实际返回张数计费。 - 每张参考图在每次请求中只计算一次,不随输出张数翻倍。例如 2 张参考图生成 3 张图片,计 2 张输入和 3 张输出。
- 生成失败的任务不收费,预扣生成费用全额退回。
常见参数错误
以下参数错误在提交时同步返回 HTTP 400,不创建任务、不扣生成费用。提交失败时读取error.message;任务创建后失败时读取 data.error.message。
| 报错片段 | 处理方式 |
|---|---|
`prompt` is required for kling-v4-image | 提供非空提示词。 |
`n` must be an integer between 1 and 9 | n 使用 1–9 的整数。 |
`n` cannot be combined with sequential_image_generation=auto | 组图的 n 可省略或设为 1,用 max_images 控制输出张数。 |
`sequential_image_generation_options.max_images` is required for kling-v4-image series mode | 开启组图时必须提供 max_images。 |
`sequential_image_generation_options.max_images` must be an integer between 2 and 4 | max_images 使用 2–4 的整数,不传 "auto"。 |
`resolution` must be 2K, 4K or 8K for kling-v4-image | 使用 2K、4K 或 8K,不区分大小写。 |
`size` must be auto or one of 16:9, 9:16, 1:1, 4:3, 3:4, 3:2, 2:3, 21:9 | 使用 auto 或支持的宽高比,不传像素尺寸。 |
kling-v4-image supports at most 10 reference images | 参考图最多 10 张。 |
reference image is not accessible for kling-v4-image | 参考图链接返回 403 / 404 / 410;检查链接是否公开且仍有效,必要时重新上传。 |
unsupported reference image dimensions for kling-v4-image | 检查宽、高均至少 300 px,且宽高比在 2:5–5:2 之间;宽高比超限使用相同报错前缀。 |
reference image too large for kling-v4-image | 单张参考图不得超过 30 MB;压缩后重新提交。 |
`n` must be an integer between 1 and 9 for kling-v4-image (got: "3") | 使用 JSON 整数,如 "n": 3,不要传字符串 "3";其他字段也须使用声明的字符串、数组或布尔类型。 |