🔥Midjourney API

Midjourney相关API，包含目前v5-v6所有操作。TTAPI 的 Midjourney API 是目前最稳定、最具成本效益的 Midjourney 产品。

模式

TT API 的 Midjourney 服务提供 relax、fast 和 turbo 三种模式的图像生成，与 Midjourney 对应。

• fast 模式响应时间在 90秒内，整体成功率 超过98%

• relax 模式响应一般在 10分钟内，根据官方响应，总体成功率超过90%

• turbo 模式响应时间在 60秒内

最多可以 同时执行 10 个生成队列，如需更多作业队列请 联系我们

生成图像[/imagine]

Midjourney 6.0 版本的 “Imagine 接口” 配额消耗将是其他模型的1.5倍。

--cref 操作必须配合 --v 6.0 模型使用

POST https://api.ttapi.io/midjourney/v1/imagine

Imagine 接口将根据文案生成4张图像。

Headers

Name	Value
TT-API-KEY	用于请求授权 TT-API 的API密钥
Content-Type	application-json

Body

参数	类型	是否必须	描述
prompt	string	是	用以生成图像的提示词 例：a cat --ar 1:1
hookUrl	string	否	回调地址，任务完成或失败将通过请地址进行通知。如果未设置，则需要请求fetch 接口进行查询。
mode	string	否	生成图像的任务模式，包括 relax、fast和turbo。如果未填写，则默认情况下将使用 fast 模式。
timeout	int	否	请求超时时间，单位：秒。如果未填写，默认超时时间 300秒
getUImages	boolean	否	是否获取四张小图，可选值为 true, false 默认为 false，对应的返回结果为JSON结构中images字段。注意：此操作并不是真正对生成的任务执行U操作，选择true之后依然可以进行U放大

如果 hookUrl 不为空，系统会向您设置的 hookUrl 发送一个 json响应。

请求示例

Python

import requests

endpoint = "https://api.ttapi.io/midjourney/v1/imagine"

headers = {
    "TT-API-KEY": your_key
}

data = {
    "prompt": "a cute cat",
    "model": "fast",
    "hookUrl": "",
    "timeout": 300
}

response = requests.post(endpoint, headers=headers, json=data)

print(response.status_code)
print(response.json())

Php

<?php

$url = 'https://api.ttapi.io/midjourney/v1/imagine';

$data = array(
    'prompt' => 'a cat --ar 1:1',
    'model' => 'fast',
    'hookUrl' => '',
    'timeout' => 300
);

$ch = curl_init($url);

$dataString = json_encode($data);

curl_setopt($ch, CURLOPT_CUSTOMREQUEST, "POST");
curl_setopt($ch, CURLOPT_POSTFIELDS, $dataString);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
curl_setopt($ch, CURLOPT_HTTPHEADER, array(
    'Content-Type: application/json',
    'TT-API-KEY: your_key'
);

$result = curl_exec($ch);

$httpCode = curl_getinfo($ch, CURLINFO_HTTP_CODE);

curl_close($ch);

echo $httpCode;

Node

const axios = require('axios');

let config = {
  method: 'post',
  url: 'https://api.ttapi.io/midjourney/v1/imagine',
  headers: { 
    'TT-API-KEY': 'your_key'
  },
  data : {
    "prompt": "a cute cat",
    "model": "fast",
    "hookUrl": "",
    "timeout": 300
  }
};

axios(config)
.then(function (response) {
  console.log(JSON.stringify(response.status));
})
.catch(function (error) {
  console.log(error);
});

响应示例

200

{
  "status": "SUCCESS",
  "message": "",
  "data": {
    "jobId": "afa774a3-1aee-5aba-4510-14818d6875e4"
  }
}

400

{
  "status": "FAILED",
  "message": "'action' parameter error!",
  "data": null
}

图像变化[U1~U4、V1~V4]

POST https://api.ttapi.io/midjourney/v1/action

此接口包含生成图像下方的按钮操作 例如：U1~U4，V1~V4等，具体可参考 Action操作

Headers

Name	Value
TT-API-KEY	用于请求授权 TT-API 的API密钥
Content-Type	application-json

Body

参数	类型	是否必须	描述
jobId	string	是	上一步请求返回的jobId
action	string	否	您的具体操作 例如： upsample1 = U1 variation1 = V1 更多请参考
timeout	int	否	请求超时时间，单位：秒。如果未填写，默认超时时间 300秒
hookUrl	string	否	任务完成或失败将通过请地址进行通知。如果未设置，则需要请求fetch 接口进行查询。
getUImages	boolean	否	是否获取四张小图，可选值为 true, false 默认为 false，对应的返回结果为JSON结构中images字段。注意：此操作并不是真正对生成的任务执行U操作，选择true之后依然可以进行U放大

请求示例

Python

import requests

endpoint = "https://api.ttapi.io/midjourney/v1/action"

headers = {
    "TT-API-KEY": your_key
}

data = {
    "jobId": "afa774a3-1aee-5aba-4510-14818d6875e4",
    "action": "upsample1",
    "hookUrl": ""
}

response = requests.post(endpoint, headers=headers, json=data)

print(response.status_code)
print(response.json())

Php

<?php

$url = 'https://api.ttapi.io/midjourney/v1/action';

$data = array(
    'jobId' => 'afa774a3-1aee-5aba-4510-14818d6875e4',
    'action' => 'upsample1',
    'hookUrl' => ''
);

$ch = curl_init($url);

$dataString = json_encode($data);

curl_setopt($ch, CURLOPT_CUSTOMREQUEST, "POST");
curl_setopt($ch, CURLOPT_POSTFIELDS, $dataString);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
curl_setopt($ch, CURLOPT_HTTPHEADER, array(
    'Content-Type: application/json',
    'TT-API-KEY: your_key'
);

$result = curl_exec($ch);

$httpCode = curl_getinfo($ch, CURLINFO_HTTP_CODE);

curl_close($ch);

echo $httpCode;

Node

const axios = require('axios');

let config = {
  method: 'post',
  url: 'https://api.ttapi.io/midjourney/v1/action',
  headers: { 
    'TT-API-KEY': 'your_key'
  },
  data : {
    "jobId": "afa774a3-1aee-5aba-4510-14818d6875e4",
    "action": "upsample1",
    "hookUrl": ""
  }
};

axios(config)
.then(function (response) {
  console.log(JSON.stringify(response.status));
})
.catch(function (error) {
  console.log(error);
});

响应示例

200

{
  "status": "SUCCESS",
  "message": "",
  "data": {
    "jobId": "afa774a3-1aee-5aba-4510-14818d6875e4"
  }
}

400

{
  "status": "FAILED",
  "message": "'action' parameter error!",
  "data": null
}

获取图像种子[--seed]

POST https://api.ttapi.io/midjourney/v1/seed

从 Midjourney 图像中获取种子，了解具体用法请参阅 Midjourney文档

Headers

Name	Value
TT-API-KEY	用于请求授权 TT-API 的API密钥
Content-Type	application-json

Body

参数	类型	是否必须	描述
jobId	string	是	上一步请求返回的jobId
timeout	int	否	请求超时时间，单位：秒。如果未填写，默认超时时间 300秒
hookUrl	string	否	任务完成或失败将通过请地址进行通知。如果未设置，则需要请求fetch 接口进行查询。

请求示例

Python

import requests

endpoint = "https://api.ttapi.io/midjourney/v1/seed"

headers = {
    "TT-API-KEY": your_key
}

data = {
    "jobId": "afa774a3-1aee-5aba-4510-14818d6875e4",
    "hookUrl": ""
}

response = requests.post(endpoint, headers=headers, json=data)

print(response.status_code)
print(response.json())

Php

<?php

$url = 'https://api.ttapi.io/midjourney/v1/seed';

$data = array(
    'jobId' => 'afa774a3-1aee-5aba-4510-14818d6875e4',
    'hookUrl' => ''
);

$ch = curl_init($url);

$dataString = json_encode($data);

curl_setopt($ch, CURLOPT_CUSTOMREQUEST, "POST");
curl_setopt($ch, CURLOPT_POSTFIELDS, $dataString);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
curl_setopt($ch, CURLOPT_HTTPHEADER, array(
    'Content-Type: application/json',
    'TT-API-KEY: your_key'
);

$result = curl_exec($ch);

$httpCode = curl_getinfo($ch, CURLINFO_HTTP_CODE);

curl_close($ch);

echo $httpCode;

Node

const axios = require('axios');

let config = {
  method: 'post',
  url: 'https://api.ttapi.io/midjourney/v1/seed',
  headers: { 
    'TT-API-KEY': 'your_key'
  },
  data : {
    "jobId": "afa774a3-1aee-5aba-4510-14818d6875e4",
    "hookUrl": ""
  }
};

axios(config)
.then(function (response) {
  console.log(JSON.stringify(response.status));
})
.catch(function (error) {
  console.log(error);
});

响应示例

200

{
  "status": "SUCCESS",
  "message": "",
  "data": {
    "jobId": "afa774a3-1aee-5aba-4510-14818d6875e4"
  }
}

图像合成[/blend]

POST https://api.ttapi.io/midjourney/v1/blend

上传2-5个图像，然后根据每个图像的概念和美学，将它们合并为一个新图像。

Headers

Name	Value
TT-API-KEY	用于请求授权 TT-API 的API密钥
Content-Type	application-json

Body

参数	类型	是否必须	描述
imgBase64Array	Array	是	将要生成混图的图片Base64数组 数组长度为 2-5 例：[ "data:image/png;base64,xxx1", "data:image/png;base64,xxx2" ]
dimensions	string	否	生成图像的比例，包括PORTRAIT 、SQUARE、LANDSCAPE。如果未填写，默认使用SQUARE。 PORTRAIT 对应比例 2：3 SQUARE 对应比例 1：1 LANDSCAPE 对应比例 3：2
mode	string	否	生成图像的任务模式，包括 relax、fast和turbo。如果未填写，则默认情况下将使用 fast 模式。
hookUrl	string	否	任务完成或失败将通过请地址进行通知。如果未设置，则需要请求fetch 接口进行查询。
timeout	int	否	请求超时时间，单位：秒。如果未填写，默认超时时间 300秒

请求示例

Python

import requests

endpoint = "https://api.ttapi.io/midjourney/v1/blend"

headers = {
    "TT-API-KEY": your_key
}

data = {
    "imgBase64Array": ["data:image/png;base64,xxx1","data:image/png;base64,xxx2"],
    "dimensions": "SQUARE",
    "model": "fast",
    "hookUrl": ""
}

response = requests.post(endpoint, headers=headers, json=data)

print(response.status_code)
print(response.json())

Php

<?php

$url = 'https://api.ttapi.io/midjourney/v1/blend';

$data = array(
    "imgBase64Array" => ["data:image/png;base64,xxx1","data:image/png;base64,xxx2"],
    "dimensions" => "SQUARE",
    "model" => "fast",
    "hookUrl" => ""
);

$ch = curl_init($url);

$dataString = json_encode($data);

curl_setopt($ch, CURLOPT_CUSTOMREQUEST, "POST");
curl_setopt($ch, CURLOPT_POSTFIELDS, $dataString);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
curl_setopt($ch, CURLOPT_HTTPHEADER, array(
    'Content-Type: application/json',
    'TT-API-KEY: your_key'
);

$result = curl_exec($ch);

$httpCode = curl_getinfo($ch, CURLINFO_HTTP_CODE);

curl_close($ch);

echo $httpCode;

Node

const axios = require('axios');

let config = {
  method: 'post',
  url: 'https://api.ttapi.io/midjourney/v1/blend',
  headers: { 
    'TT-API-KEY': 'your_key'
  },
  data : {
    "imgBase64Array": ["data:image/png;base64,xxx1","data:image/png;base64,xxx2"],
    "dimensions": "SQUARE",
    "model": "fast",
    "hookUrl": ""
  }
};

axios(config)
.then(function (response) {
  console.log(JSON.stringify(response.status));
})
.catch(function (error) {
  console.log(error);
});

响应示例

200

{
  "status": "SUCCESS",
  "message": "",
  "data": {
    "jobId": "afa774a3-1aee-5aba-4510-14818d6875e4"
  }
}

图像描述词[/describe]

POST https://api.ttapi.io/midjourney/v1/describe

上传图像，并根据该图像生成四个提示信息。

Headers

Name	Value
TT-API-KEY	用于请求授权 TT-API 的API密钥
Content-Type	application-json

Body

参数	类型	是否必须	描述
base64	string	是	图像的base64编码 例："data:image/png;base64,xxx1"
mode	string	否	生成图像的任务模式，包括 relax、fast和turbo。如果未填写，则默认情况下将使用 fast 模式。
hookUrl	string	否	任务完成或失败将通过请地址进行通知。如果未设置，则需要请求fetch 接口进行查询。
timeout	int	否	请求超时时间，单位：秒。如果未填写，默认超时时间 300秒

请求示例

Python

import requests

endpoint = "https://api.ttapi.io/midjourney/v1/describe"

headers = {
    "TT-API-KEY": your_key
}

data = {
    "base64": "data:image/png;base64,xxx1",
    "model": "fast",
    "hookUrl": ""
}

response = requests.post(endpoint, headers=headers, json=data)

print(response.status_code)
print(response.json())

Php

<?php

$url = 'https://api.ttapi.io/midjourney/v1/describe';

$data = array(
    "base64" => "data:image/png;base64,xxx1",
    "model" => "fast",
    "hookUrl" => ""
);

$ch = curl_init($url);

$dataString = json_encode($data);

curl_setopt($ch, CURLOPT_CUSTOMREQUEST, "POST");
curl_setopt($ch, CURLOPT_POSTFIELDS, $dataString);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
curl_setopt($ch, CURLOPT_HTTPHEADER, array(
    'Content-Type: application/json',
    'TT-API-KEY: your_key'
);

$result = curl_exec($ch);

$httpCode = curl_getinfo($ch, CURLINFO_HTTP_CODE);

curl_close($ch);

echo $httpCode;

Node

const axios = require('axios');

let config = {
  method: 'post',
  url: 'https://api.ttapi.io/midjourney/v1/describe',
  headers: { 
    'TT-API-KEY': 'your_key'
  },
  data : {
    "base64": "data:image/png;base64,xxx1",
    "model": "fast",
    "hookUrl": ""
  }
};

axios(config)
.then(function (response) {
  console.log(JSON.stringify(response.status));
})
.catch(function (error) {
  console.log(error);
});

响应示例

200

{
  "status": "SUCCESS",
  "message": "",
  "data": {
    "jobId": "afa774a3-1aee-5aba-4510-14818d6875e4"
  }
}

区域重绘[Vary(region)]

POST https://api.ttapi.io/midjourney/v1/inpaint

区域重绘，等同midjourney的Vary(region)按钮

Headers

Name	Value
TT-API-KEY	用于请求授权 TT-API 的API密钥
Content-Type	application-json

Body

参数	类型	是否必须	描述
jobId	string	是	上一步请求返回的jobId
mask	string	是	所选择图片的Base64值
prompt	string	否	对所选择图片区域想要修改的描述
timeout	int	否	请求超时时间，单位：秒。如果未填写，默认超时时间 300秒
hookUrl	string	否	任务完成或失败将通过请地址进行通知。如果未设置，则需要请求fetch 接口进行查询。

请求示例

Python

import requests

endpoint = "https://api.ttapi.io/midjourney/v1/inpaint"

headers = {
    "TT-API-KEY": your_key
}

data = {
   "jobId": "afa774a3-1aee-5aba-4510-14818d6875e4",
    "prompt": "white background",
    "mask": "UklGRrw0AABXRUJQVlA4WAoAAAAgAAAA..."
    "hookUrl": ""
}

response = requests.post(endpoint, headers=headers, json=data)

print(response.status_code)
print(response.json())

Php

<?php

$url = 'https://api.ttapi.io/midjourney/v1/inpaint';

$data = array(
    "jobId" => "afa774a3-1aee-5aba-4510-14818d6875e4",
    "prompt" => "white background",
    "mask"   => "UklGRrw0AABXRUJQVlA4WAoAAAAgAAAA..."
    "hookUrl" => ""
);

$ch = curl_init($url);

$dataString = json_encode($data);

curl_setopt($ch, CURLOPT_CUSTOMREQUEST, "POST");
curl_setopt($ch, CURLOPT_POSTFIELDS, $dataString);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
curl_setopt($ch, CURLOPT_HTTPHEADER, array(
    'Content-Type: application/json',
    'TT-API-KEY: your_key'
);

$result = curl_exec($ch);

$httpCode = curl_getinfo($ch, CURLINFO_HTTP_CODE);

curl_close($ch);

echo $httpCode;

Node

const axios = require('axios');

let config = {
  method: 'post',
  url: 'https://api.ttapi.io/midjourney/v1/inpaint',
  headers: { 
    'TT-API-KEY': 'your_key'
  },
  data : {
    "jobId" => "afa774a3-1aee-5aba-4510-14818d6875e4",
    "prompt" => "white background",
    "mask"   => "UklGRrw0AABXRUJQVlA4WAoAAAAgAAAA..."
    "hookUrl" => ""
  }
};

axios(config)
.then(function (response) {
  console.log(JSON.stringify(response.status));
})
.catch(function (error) {
  console.log(error);
});

响应示例

200

{
  "status": "SUCCESS",
  "message": "",
  "data": {
    "jobId": "afa774a3-1aee-5aba-4510-14818d6875e4"
  }
}

400

{
  "status": "FAILED",
  "message": "'action' parameter error!",
  "data": {}
}

获取任务状态 - fetch

POST https://api.ttapi.io/midjourney/v1/fetch

查询任务结果

Headers

Name	Value
TT-API-KEY	用于请求授权 TT-API 的API密钥
Content-Type	application-json

Body

参数	类型	是否必须	描述
jobId	string	是	上一步请求返回的jobId

Response

返回结果与异步回调数据结构一致

Action 具体使用操作

关于相关操作的具体用法，您可以详细阅读 Midjourney官方文档

操作名称	描述	midjourney 示例
upsample1	U1-U4 按钮 upsample2, upsample3, upsample4 同理	upsample
variation1	V1-V4 按钮 variation2, variation3, variation4 同理	variation
high_variation	Vary (Strong)按钮 对图片进行大幅变化	high_variation
low_variation	Vary (Subtle) 按钮 对图片进行小幅变化	low_variation
upscale2	Upscale(2x) 按钮 图像放大 2 倍，只存在于v5模式	upscale2
upscale4	Upscale(4x) 按钮 图像放大 4 倍，只存在于v5模式	upscale4
zoom_out_2	Zoom Out 2x 按钮 “缩小”选项在不更改原始图片的情况下展开放大图像的画布，根据提示和初始图像填充新空间。	zoom_out_2
zoom_out_1_5	Zoom Out 1.5x 按钮 “缩小”选项在不更改原始图片的情况下展开放大图像的画布，根据提示和初始图像填充新空间。	zoom_out_1_5
pan_left	⬅️➡️⬆️⬇️ 按钮 在选定方向上平移展开图像画布，而不会更改原始内容。根据初始图像和提供的提示填充扩展区域。pan_right, pan_up, pan_down 同理	pan.png
upscale_creative	Upscale(Creative) 按钮 只存在于v6模式	upscale_creative.png
upscale_subtle	Upscale(Subtle) 按钮 只存在于v6模式	upscale_subtle.png

异步回调 JSON 结构

SUCCESS

{
    "status": "SUCCESS",
    "jobId": "f5850038-90a3-8a97-0476-107ea4b8dac4",
    "message": "success",
    "data": {
        "actions": "imagine",
        "jobId": "f5850038-90a3-8a97-0476-107ea4b8dac4",
        "progress": "100",
        "prompt": "Soccer star Max Kruse and Jan-Peter Jachtmann victims of €528,695 poker scam, German soccer star Max Kruse and WSOP Main Event finalist Jan-Peter Jachtmann are among the players who have been swindled out of €528,695., poker, realistic --ar 1280:720",
        "discordImage": "https://cdn.discordapp.com/attachments/1107938555931656214/1176340921227423844/voyagel_Soccer_star_Max_Kruse_and_Jan-Peter_Jachtmann_victims_o_c513a87b-eed3-4a3b-ab97-6be4dbc3ea99.png?ex=656e83da&is=655c0eda&hm=6e06a1dec3c6c1be209799884681969878eabb81ce81f8db22d54480379fcd9b&",
        "cdnImage": "http://127.0.0.1/8080/pics/452020f2-6793-4525-a1b5-472cac439610.png",
        "hookUrl": "",
        "components": [
            "upsample1",
            "upsample2",
            "upsample3",
            "upsample4",
            "variation1",
            "variation2",
            "variation3",
            "variation4"
        ],
        "seed":"",
        "images": [
            "https://cdnb.ttapi.io/2024-04-02/5ac3b82974cfbc394ff70fdb3f8ad243a78cce8c3cd88defee12739db0b499f9.png",
            "https://cdnb.ttapi.io/2024-04-02/41c7a7dfc8ec9871d7aa6f45bbcb3ca7dcb967cadef285ae3b05b2fce63e9fa5.png",
            "https://cdnb.ttapi.io/2024-04-02/1d62e5110d24de42e8ac4894108874885b8460e5cbd212fe19ad756b1bf15087.png",
            "https://cdnb.ttapi.io/2024-04-02/a1887b8f816286df67cc83d0b1cc07450b2558bc28a60fbea4991f4ae390652c.png"
        ]
    }
}

PENDING_QUEUE

{
    "status": "PENDING_QUEUE",
    "jobId": "f5850038-90a3-8a97-0476-107ea4b8dac4",
    "message": "",
    "data": []
}

ON_QUEUE

{
    "status": "ON_QUEUE",
    "jobId": "f5850038-90a3-8a97-0476-107ea4b8dac4",
    "message": "",
    "data": {
        "actions": "imagine",
        "jobId": "f5850038-90a3-8a97-0476-107ea4b8dac4",
        "progress": "100",
        "prompt": "Soccer star Max Kruse and Jan-Peter Jachtmann victims of €528,695 poker scam, German soccer star Max Kruse and WSOP Main Event finalist Jan-Peter Jachtmann are among the players who have been swindled out of €528,695., poker, realistic --ar 1280:720",
        "discordImage": "https://cdn.discordapp.com/attachments/1107938555931656214/1176340921227423844/voyagel_Soccer_star_Max_Kruse_and_Jan-Peter_Jachtmann_victims_o_c513a87b-eed3-4a3b-ab97-6be4dbc3ea99.png?ex=656e83da&is=655c0eda&hm=6e06a1dec3c6c1be209799884681969878eabb81ce81f8db22d54480379fcd9b&",
        "cdnImage": "http://127.0.0.1/8080/pics/452020f2-6793-4525-a1b5-472cac439610.png",
        "hookUrl": "",
        "components": [],
        "seed": null,
        "images": null
    }
}

FAILED

{
    "status": "FAILED",
    "jobId": "f5850038-90a3-8a97-0476-107ea4b8dac4",
    "message": "failed",
    "data": []
}

从2023年12月1日开始，discord返回的cdn图片链接将在24-72小时之内过期。

返回JSON结构详解

name	value
status	PENDING_QUEUE //排队中 ON_QUEUE //执行中 SUCCESS //成功 FAILED //失败
progess	进度 : 0-100，100为完成
components	支持的下一步操作
discordImage	discord的cdn图片地址
cdnImage	国内可访问的cdn代理地址
images	ttapi的cdn地址，imagine命令生成的四张小图