Merge branch 'develop' of https://github.com/HBAI-Ltd/Toonflow-app into develop
This commit is contained in:
小帅
commit
e5d860888b
@ -51,10 +51,9 @@ export default router.post(
|
||||
console.log("%c Line:52 🍐 reply", "background:#ffdd4d", reply);
|
||||
res.status(200).send(success(reply));
|
||||
} catch (err) {
|
||||
console.log(err);
|
||||
if (typeof err === "string") return res.status(500).send(error(err));
|
||||
const msg = err instanceof Error ? err.message : (err as any)?.error?.message;
|
||||
return res.status(500).send(error(msg || "未知错误"));
|
||||
const msg = u.error(err).message;
|
||||
console.error(msg);
|
||||
res.status(500).send(error(msg));
|
||||
}
|
||||
},
|
||||
);
|
||||
|
||||
@ -18,36 +18,17 @@ export default router.post(
|
||||
const { modelName, apiKey, baseURL, manufacturer } = req.body;
|
||||
try {
|
||||
const image = await u.ai.image({
|
||||
prompt: "生成16：9 四宫格图片，第一宫格是一只猫，第二宫格是一只狗， 第三宫格是一只老虎，第四宫格是猪。保证四宫格图片标准四等分",
|
||||
prompt:
|
||||
"一张16:9比例的图片，完美等分为2x2四宫格布局，各区域无缝衔接：\n左上宫格：一只可爱的猫，毛发蓬松，眼睛明亮，姿态俏皮\n右上宫格：一只友善的狗，金毛犬，表情愉悦，摇着尾巴\n左下宫格：一头健壮的牛，田园背景，目光温和，皮毛光泽\n右下宫格：一匹骏马，姿态优雅，鬃毛飘逸，肌肉健美\n风格要求：四个宫格风格统一，色彩鲜艳饱和，高清画质，细节清晰锐利，专业插画风格，线条干净，统一的左上方光源，柔和阴影，和谐配色，卡通/半写实风格，宫格间用白色或浅灰细线分隔",
|
||||
imageBase64: [],
|
||||
aspectRatio: "16:9",
|
||||
size: "1K",
|
||||
});
|
||||
res.status(200).send(success(image));
|
||||
} catch (e: any) {
|
||||
console.log("%c Line:28 🥒 e", "background:#fca650", e);
|
||||
return res.status(500).send(error(e?.response?.data ?? e?.message ?? "生成失败"));
|
||||
} catch (err) {
|
||||
const msg = u.error(err).message;
|
||||
console.error(msg);
|
||||
res.status(500).send(error(msg));
|
||||
}
|
||||
|
||||
// try {
|
||||
// const contentStr = await u.ai.generateImage(
|
||||
// {
|
||||
// prompt: "2D cat",
|
||||
// imageBase64: [],
|
||||
// aspectRatio: "16:9",
|
||||
// size: "1K",
|
||||
// },
|
||||
// {
|
||||
// model: modelName,
|
||||
// apiKey,
|
||||
// baseURL,
|
||||
// manufacturer,
|
||||
// },
|
||||
// );
|
||||
// res.status(200).send(success(contentStr));
|
||||
// } catch (err: any) {
|
||||
// const message = err?.response?.data?.error?.message || err?.error?.message || "模型调用失败";
|
||||
// res.status(500).send(error(message));
|
||||
// }
|
||||
},
|
||||
);
|
||||
|
||||
@ -20,20 +20,21 @@ export default router.post(
|
||||
async (req, res) => {
|
||||
const { modelName, apiKey, baseURL, manufacturer } = req.body;
|
||||
try {
|
||||
const videoPath = await u.ai.generateVideo(
|
||||
{
|
||||
imageBase64: [],
|
||||
savePath: "",
|
||||
prompt: "stickman Dances",
|
||||
duration: 10 as any,
|
||||
aspectRatio: "16:9" as any,
|
||||
},
|
||||
manufacturer,
|
||||
);
|
||||
const videoPath = await u.ai.video({
|
||||
imageBase64: [],
|
||||
savePath: "test.mp4",
|
||||
prompt: "stickman Dances",
|
||||
duration: 4,
|
||||
resolution: "720p",
|
||||
aspectRatio: "16:9",
|
||||
audio: false,
|
||||
});
|
||||
const url = await u.oss.getFileUrl(videoPath);
|
||||
res.status(200).send(success(url));
|
||||
} catch (err: any) {
|
||||
res.status(500).send(error(err.error.message || "模型调用失败"));
|
||||
const msg = u.error(err).message;
|
||||
console.error(msg);
|
||||
res.status(500).send(error(msg));
|
||||
}
|
||||
},
|
||||
);
|
||||
|
||||
@ -11,6 +11,7 @@ import * as imageTools from "@/utils/imageTools";
|
||||
|
||||
import AIText from "@/utils/ai/text/index";
|
||||
import AIImage from "@/utils/ai/image/index";
|
||||
import AIVideo from "@/utils/ai/video/index";
|
||||
|
||||
export default {
|
||||
db,
|
||||
@ -18,6 +19,7 @@ export default {
|
||||
ai: {
|
||||
text: AIText,
|
||||
image: AIImage,
|
||||
video: AIVideo,
|
||||
},
|
||||
editImage,
|
||||
number2Chinese,
|
||||
|
||||
561
src/utils/ai.ts
561
src/utils/ai.ts
@ -1,561 +0,0 @@
|
||||
import axios from "axios";
|
||||
import u from "@/utils";
|
||||
import FormData from "form-data";
|
||||
import axiosRetry from "axios-retry";
|
||||
import { OpenAIChatModel, type OpenAIChatModelOptions } from "@aigne/openai";
|
||||
import sharp from "sharp";
|
||||
|
||||
axiosRetry(axios, { retries: 3, retryDelay: () => 200 });
|
||||
|
||||
export const text = async (config: OpenAIChatModelOptions = {}) => {
|
||||
const { model, apiKey, baseURL } = await u.getConfig("text");
|
||||
return new OpenAIChatModel({
|
||||
apiKey: apiKey ?? "",
|
||||
baseURL: baseURL ?? "",
|
||||
model: model ?? "gpt-4.1",
|
||||
modelOptions: { temperature: 0.7 },
|
||||
...config,
|
||||
});
|
||||
};
|
||||
|
||||
interface ImageConfig {
|
||||
systemPrompt?: string;
|
||||
prompt: string;
|
||||
imageBase64: string[];
|
||||
size: "1K" | "2K" | "4K";
|
||||
aspectRatio: string;
|
||||
resType?: "url" | "b64";
|
||||
}
|
||||
|
||||
const urlToBase64 = async (url: string): Promise<string> => {
|
||||
const res = await axios.get(url, { responseType: "arraybuffer" });
|
||||
const base64 = Buffer.from(res.data).toString("base64");
|
||||
const mimeType = res.headers["content-type"] || "image/png";
|
||||
return `data:${mimeType};base64,${base64}`;
|
||||
};
|
||||
|
||||
const sleep = (ms: number) => new Promise((resolve) => setTimeout(resolve, ms));
|
||||
|
||||
const pollTask = async (
|
||||
queryFn: () => Promise<{ completed: boolean; imageUrl?: string; error?: string }>,
|
||||
maxAttempts = 500,
|
||||
interval = 2000,
|
||||
): Promise<string> => {
|
||||
for (let i = 0; i < maxAttempts; i++) {
|
||||
await sleep(interval);
|
||||
const { completed, imageUrl, error } = await queryFn();
|
||||
if (error) throw new Error(error);
|
||||
if (completed && imageUrl) return imageUrl;
|
||||
}
|
||||
throw new Error(`任务轮询超时，已尝试 ${maxAttempts} 次`);
|
||||
};
|
||||
|
||||
// 上传 base64 图片到 runninghub
|
||||
const uploadBase64ToRunninghub = async (base64Image: string, apiKey: string, baseURL: string): Promise<string> => {
|
||||
try {
|
||||
apiKey = apiKey.replace("Bearer ", "");
|
||||
// 移除 base64 前缀
|
||||
const base64Data = base64Image.replace(/^data:image\/\w+;base64,/, "");
|
||||
let buffer = Buffer.from(base64Data, "base64");
|
||||
|
||||
// 压缩图片到 5MB 以下
|
||||
const MAX_SIZE = 5 * 1024 * 1024; // 5MB
|
||||
if (buffer.length > MAX_SIZE) {
|
||||
let quality = 90;
|
||||
|
||||
while (buffer.length > MAX_SIZE && quality > 10) {
|
||||
const compressed = await sharp(buffer).jpeg({ quality, mozjpeg: true }).toBuffer();
|
||||
buffer = Buffer.from(compressed);
|
||||
quality -= 10;
|
||||
}
|
||||
|
||||
// 如果仍然超过限制，进一步调整尺寸
|
||||
if (buffer.length > MAX_SIZE) {
|
||||
const metadata = await sharp(buffer).metadata();
|
||||
const scale = Math.sqrt(MAX_SIZE / buffer.length);
|
||||
|
||||
const resized = await sharp(buffer)
|
||||
.resize({
|
||||
width: Math.floor((metadata.width || 1920) * scale),
|
||||
height: Math.floor((metadata.height || 1080) * scale),
|
||||
fit: "inside",
|
||||
})
|
||||
.jpeg({ quality: 80, mozjpeg: true })
|
||||
.toBuffer();
|
||||
|
||||
buffer = Buffer.from(resized);
|
||||
}
|
||||
}
|
||||
|
||||
// 创建 FormData
|
||||
const formData = new FormData();
|
||||
formData.append("file", buffer, {
|
||||
filename: "image.jpg",
|
||||
contentType: "image/jpeg",
|
||||
});
|
||||
|
||||
// 上传图片
|
||||
const uploadRes = await axios.post(`https://www.runninghub.cn/openapi/v2/media/upload/binary`, formData, {
|
||||
headers: { Authorization: `Bearer ${apiKey}` },
|
||||
});
|
||||
|
||||
if (uploadRes.data.code !== 0 || !uploadRes.data.data?.download_url) {
|
||||
throw new Error(`图片上传失败: ${JSON.stringify(uploadRes.data)}`);
|
||||
}
|
||||
|
||||
return uploadRes.data.data.download_url;
|
||||
} catch (error) {
|
||||
console.error("上传图片时发生错误:", error);
|
||||
throw error;
|
||||
}
|
||||
};
|
||||
|
||||
const generators = {
|
||||
volcengine: async (config: ImageConfig, apiKey: string, baseURL: string, model: string) => {
|
||||
if (config.size == "1K") config.size = "2K";
|
||||
apiKey = apiKey.replace("Bearer ", "");
|
||||
const body: Record<string, any> = {
|
||||
model,
|
||||
prompt: config.prompt,
|
||||
size: config.size,
|
||||
response_format: "url",
|
||||
sequential_image_generation: "disabled",
|
||||
stream: false,
|
||||
watermark: false,
|
||||
};
|
||||
// 图生图：存在图片时添加 image 字段
|
||||
if (config.imageBase64) {
|
||||
body.image = config.imageBase64;
|
||||
}
|
||||
const res = await axios.post(`https://ark.cn-beijing.volces.com/api/v3/images/generations`, body, {
|
||||
headers: { Authorization: `Bearer ${apiKey}` },
|
||||
});
|
||||
return res.data.data[0].url;
|
||||
},
|
||||
|
||||
gemini: async (config: ImageConfig, apiKey: string, baseURL: string, model: string) => {
|
||||
apiKey = apiKey.replace("Bearer ", "");
|
||||
const messages = [
|
||||
...(config.systemPrompt ? [{ role: "system", content: config.systemPrompt }] : []),
|
||||
{ role: "user", content: config.prompt },
|
||||
...config.imageBase64.map((img) => ({ role: "user", content: { image: img } })),
|
||||
];
|
||||
const res = await axios.post(
|
||||
`${baseURL}/chat/completions`,
|
||||
{ model, stream: false, messages, extra_body: { google: { image_config: { aspect_ratio: config.aspectRatio, image_size: config.size } } } },
|
||||
{ headers: { Authorization: "Bearer " + apiKey } },
|
||||
);
|
||||
|
||||
return res.data.choices[0].message.content;
|
||||
},
|
||||
|
||||
runninghub: async (config: ImageConfig, apiKey: string, baseURL: string) => {
|
||||
apiKey = apiKey.replace("Bearer ", "");
|
||||
const imageUrls = await Promise.all(config.imageBase64.map((base64Image) => uploadBase64ToRunninghub(base64Image, apiKey, baseURL)));
|
||||
|
||||
const endpoint = config.imageBase64.length === 0 ? "/openapi/v2/rhart-image-n-pro/text-to-image" : "/openapi/v2/rhart-image-n-pro/edit";
|
||||
const taskRes = await axios.post(
|
||||
`https://www.runninghub.cn${endpoint}`,
|
||||
{ prompt: config.prompt, resolution: config.size, aspectRatio: config.aspectRatio, ...(imageUrls.length > 0 && { imageUrls }) },
|
||||
{ headers: { Authorization: "Bearer " + apiKey } },
|
||||
);
|
||||
const taskId = taskRes.data.taskId;
|
||||
if (!taskId) throw new Error(`任务创建失败，${JSON.stringify(taskRes.data)}`);
|
||||
|
||||
return pollTask(async () => {
|
||||
const res = await axios.post(`https://www.runninghub.cn/task/openapi/outputs`, { taskId, apiKey: apiKey });
|
||||
const { code, msg, data } = res.data;
|
||||
if (code === 0 && msg === "success") return { completed: true, imageUrl: data?.[0]?.fileUrl };
|
||||
if (code === 804 || code === 813) return { completed: false };
|
||||
if (code === 805) return { completed: false, error: `任务失败: ${data?.[0]?.failedReason?.exception_message || "未知原因"}` };
|
||||
return { completed: false, error: `未知状态: code=${code}, msg=${msg}` };
|
||||
});
|
||||
},
|
||||
|
||||
apimart: async (config: ImageConfig, apiKey: string, baseURL: string, model: string) => {
|
||||
apiKey = apiKey.replace("Bearer ", "");
|
||||
const taskRes = await axios.post(
|
||||
`https://api.apimart.ai/v1/images/generations`,
|
||||
{ model: "gemini-3-pro-image-preview", prompt: config.prompt, size: config.aspectRatio, n: 1, resolution: config.size },
|
||||
{ headers: { Authorization: apiKey } },
|
||||
);
|
||||
|
||||
if (taskRes.data.code !== 200 || !taskRes.data.data?.[0]?.task_id) throw new Error("任务创建失败: " + JSON.stringify(taskRes.data));
|
||||
|
||||
const taskId = taskRes.data.data[0].task_id;
|
||||
return pollTask(async () => {
|
||||
const res = await axios.get(`https://api.apimart.ai/v1/tasks/${taskId}`, { headers: { Authorization: apiKey }, params: { language: "en" } });
|
||||
if (res.data.code !== 200) return { completed: false, error: `查询失败: ${JSON.stringify(res.data)}` };
|
||||
const { status, result } = res.data.data;
|
||||
if (status === "completed") return { completed: true, imageUrl: result?.images?.[0]?.url?.[0] };
|
||||
if (status === "failed" || status === "cancelled") return { completed: false, error: `任务${status}` };
|
||||
return { completed: false };
|
||||
});
|
||||
},
|
||||
};
|
||||
|
||||
export const generateImage = async (config: ImageConfig, replaceConfig?: Awaited<ReturnType<typeof u.getConfig<"image">>>): Promise<string> => {
|
||||
let { model, apiKey, baseURL, manufacturer } = await u.getConfig("image");
|
||||
if (replaceConfig) {
|
||||
model = replaceConfig.model || model;
|
||||
apiKey = replaceConfig.apiKey || apiKey;
|
||||
baseURL = replaceConfig.baseURL || baseURL;
|
||||
manufacturer = replaceConfig.manufacturer || manufacturer;
|
||||
}
|
||||
const generator = generators[manufacturer as keyof typeof generators];
|
||||
if (!generator) throw new Error(`不支持的厂商: ${manufacturer}`);
|
||||
|
||||
let imageUrl = await generator(config, apiKey ?? "", baseURL ?? "", model ?? "");
|
||||
if (!config.resType) config.resType = "b64";
|
||||
if (config.resType === "b64" && imageUrl.startsWith("http")) imageUrl = await urlToBase64(imageUrl);
|
||||
return imageUrl;
|
||||
};
|
||||
|
||||
type VideoAspectRatio = "16:9" | "9:16" | "1:1" | "4:3" | "3:4" | "21:9" | "adaptive";
|
||||
interface BaseVideoConfig {
|
||||
prompt: string;
|
||||
savePath: string;
|
||||
imageBase64?: string[]; // 单张参考图片 base64
|
||||
}
|
||||
interface DoubaoVideoConfig extends BaseVideoConfig {
|
||||
duration: 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12; // 支持 2~12 秒
|
||||
aspectRatio: VideoAspectRatio;
|
||||
audio?: boolean;
|
||||
}
|
||||
interface RunninghubVideoConfig extends BaseVideoConfig {
|
||||
duration: 10 | 15; // 仅支持 10 或 15 秒
|
||||
aspectRatio: "16:9" | "9:16" | "1:1"; // 仅支持这三种比例
|
||||
}
|
||||
interface OpenAIVideoConfig extends BaseVideoConfig {
|
||||
duration: 10 | 15; // 仅支持 10 或 15 秒
|
||||
aspectRatio: Exclude<VideoAspectRatio, "adaptive">; // 不支持 adaptive
|
||||
}
|
||||
type VideoConfig = DoubaoVideoConfig | RunninghubVideoConfig | OpenAIVideoConfig;
|
||||
const generateVideoWithConfig = async (config: VideoConfig, configItem: { model: string; apiKey: string; baseURL: string; manufacturer: string }) => {
|
||||
const { apiKey, baseURL, manufacturer, model } = configItem;
|
||||
const imageArrPath = [];
|
||||
for (const imageVal of config?.imageBase64!) {
|
||||
// 判断是否为base64串
|
||||
const isBase64 = typeof imageVal === "string" && /^data:image\/[a-zA-Z0-9\+\-\.]+;base64,[\s\S]+$/.test(imageVal.trim());
|
||||
if (isBase64) {
|
||||
imageArrPath.push(imageVal);
|
||||
} else {
|
||||
const base64 = await urlToBase64(imageVal);
|
||||
imageArrPath.push(base64);
|
||||
}
|
||||
}
|
||||
config.imageBase64 = imageArrPath;
|
||||
let videoUrl: string | null = null;
|
||||
if (manufacturer === "volcengine") {
|
||||
const doubaoConfig = config as DoubaoVideoConfig;
|
||||
const createRes = await axios.post(
|
||||
baseURL ?? "https://ark.cn-beijing.volces.com/api/v3/contents/generations/tasks",
|
||||
{
|
||||
model: "doubao-seedance-1-5-pro-251215",
|
||||
content: [
|
||||
{ type: "text", text: config.prompt },
|
||||
...(doubaoConfig.imageBase64
|
||||
? doubaoConfig.imageBase64.map((base64, i) => ({
|
||||
type: "image_url",
|
||||
image_url: { url: base64 },
|
||||
role: i === 0 ? "first_frame" : "last_frame",
|
||||
}))
|
||||
: []),
|
||||
],
|
||||
generate_audio: doubaoConfig.audio ?? false,
|
||||
duration: doubaoConfig.duration,
|
||||
resolution: doubaoConfig.aspectRatio,
|
||||
watermark: false,
|
||||
},
|
||||
{ headers: { "Content-Type": "application/json", Authorization: `Bearer ${apiKey}` } },
|
||||
);
|
||||
const taskId = createRes.data.id;
|
||||
if (!taskId) throw new Error("视频任务创建失败");
|
||||
videoUrl = await pollTask(async () => {
|
||||
const res = await axios.get(`${baseURL ?? "https://ark.cn-beijing.volces.com/api/v3/contents/generations/tasks"}/${taskId}`, {
|
||||
headers: { Authorization: `Bearer ${apiKey}` },
|
||||
});
|
||||
const { status, content } = res.data;
|
||||
if (status === "succeeded") return { completed: true, imageUrl: content?.video_url };
|
||||
if (["failed", "cancelled", "expired"].includes(status)) return { completed: false, error: `任务${status}` };
|
||||
if (["queued", "running"].includes(status)) return { completed: false };
|
||||
return { completed: false, error: `未知状态: ${status}` };
|
||||
});
|
||||
} else if (manufacturer === "runninghub") {
|
||||
const runninghubConfig = config as RunninghubVideoConfig;
|
||||
// 如果有图片，先上传
|
||||
let uploadedImageUrl: string | undefined;
|
||||
if (runninghubConfig.imageBase64 && runninghubConfig.imageBase64.length > 0) {
|
||||
uploadedImageUrl = await uploadBase64ToRunninghub(runninghubConfig.imageBase64[0]!, apiKey ?? "", "https://www.runninghub.cn");
|
||||
}
|
||||
|
||||
const endpoint = uploadedImageUrl ? "/openapi/v2/rhart-video-s/image-to-video" : "/openapi/v2/rhart-video-s/text-to-video";
|
||||
const requestBody = uploadedImageUrl
|
||||
? {
|
||||
prompt: config.prompt,
|
||||
imageUrl: uploadedImageUrl,
|
||||
duration: String(runninghubConfig.duration) as "10" | "15",
|
||||
aspectRatio: runninghubConfig.aspectRatio,
|
||||
}
|
||||
: { prompt: config.prompt, model };
|
||||
const createRes = await axios.post(`https://www.runninghub.cn${endpoint}`, requestBody, {
|
||||
headers: { Authorization: "Bearer " + apiKey, "Content-Type": "application/json" },
|
||||
});
|
||||
|
||||
const { taskId, status: initialStatus, errorMessage } = createRes.data;
|
||||
if (!taskId) throw new Error(`视频任务创建失败: ${errorMessage || "未知错误"}`);
|
||||
if (initialStatus === "FAILED") throw new Error(`任务创建失败: ${errorMessage}`);
|
||||
videoUrl = await pollTask(async () => {
|
||||
const res = await axios.post(
|
||||
`https://www.runninghub.cn/task/openapi/outputs`,
|
||||
{ apiKey: apiKey?.replace("Bearer ", ""), taskId },
|
||||
{ headers: { Authorization: "Bearer " + apiKey } },
|
||||
);
|
||||
const { code, msg, data } = res.data;
|
||||
|
||||
// 成功完成
|
||||
if (code === 0 && msg === "success" && data?.[0]?.fileUrl) {
|
||||
return { completed: true, imageUrl: data[0].fileUrl };
|
||||
}
|
||||
|
||||
// 进行中
|
||||
if (code === 804 || code === 813) {
|
||||
return { completed: false };
|
||||
}
|
||||
|
||||
// 失败
|
||||
if (code === 805) {
|
||||
const failedReason = data?.[0]?.failedReason;
|
||||
let errorMsg = "未知原因";
|
||||
|
||||
if (failedReason) {
|
||||
// 尝试多种可能的错误信息字段
|
||||
errorMsg =
|
||||
failedReason.exception_message ||
|
||||
failedReason.exceptionMessage ||
|
||||
failedReason.message ||
|
||||
failedReason.reason ||
|
||||
JSON.stringify(failedReason);
|
||||
}
|
||||
|
||||
return {
|
||||
completed: false,
|
||||
error: `任务失败: ${errorMsg}`,
|
||||
};
|
||||
}
|
||||
|
||||
// 其他未知状态
|
||||
return {
|
||||
completed: false,
|
||||
error: `未知状态: code=${code}, msg=${msg}, data=${JSON.stringify(data)}`,
|
||||
};
|
||||
});
|
||||
} else if (manufacturer === "openAi") {
|
||||
const openaiConfig = config as OpenAIVideoConfig;
|
||||
// 如果有图片，先上传
|
||||
let uploadedImageUrl: string | undefined;
|
||||
if (openaiConfig.imageBase64 && openaiConfig.imageBase64.length) {
|
||||
const base64Data = openaiConfig.imageBase64[0]!.replace(/^data:image\/\w+;base64,/, "");
|
||||
const buffer = Buffer.from(base64Data, "base64");
|
||||
const formData = new FormData();
|
||||
formData.append("file", buffer, { filename: "image.jpg", contentType: "image/jpeg" });
|
||||
const uploadRes = await axios.post(`${baseURL}/videos`, formData, {
|
||||
headers: {
|
||||
Authorization: `Bearer ${apiKey}`,
|
||||
...formData.getHeaders(),
|
||||
},
|
||||
});
|
||||
uploadedImageUrl = uploadRes.data?.id || uploadRes.data?.url;
|
||||
}
|
||||
|
||||
// 创建视频生成任务
|
||||
const formData = new FormData();
|
||||
formData.append("model", model);
|
||||
formData.append("prompt", config.prompt);
|
||||
formData.append("seconds", String(openaiConfig.duration));
|
||||
|
||||
// 根据 aspectRatio 设置 size
|
||||
const sizeMap: Record<string, string> = {
|
||||
"16:9": "1920x1080",
|
||||
"9:16": "1080x1920",
|
||||
"1:1": "1080x1080",
|
||||
"4:3": "1440x1080",
|
||||
"3:4": "1080x1440",
|
||||
"21:9": "2560x1080",
|
||||
};
|
||||
formData.append("size", sizeMap[openaiConfig.aspectRatio] || "1920x1080");
|
||||
if (uploadedImageUrl) {
|
||||
formData.append("input_reference", uploadedImageUrl);
|
||||
}
|
||||
const createRes = await axios.post(`${baseURL}/videos`, formData, {
|
||||
headers: {
|
||||
Authorization: `Bearer ${apiKey}`,
|
||||
...formData.getHeaders(),
|
||||
},
|
||||
});
|
||||
|
||||
const taskId = createRes.data?.id;
|
||||
|
||||
if (!taskId) throw new Error("视频任务创建失败");
|
||||
// 轮询任务状态
|
||||
videoUrl = await pollTask(async () => {
|
||||
const res = await axios.get(`${baseURL}/videos/${taskId}`, {
|
||||
headers: {
|
||||
Authorization: `Bearer ${apiKey}`,
|
||||
"Content-Type": "application/json",
|
||||
},
|
||||
});
|
||||
const { status, imageUrl, failReason } = res.data;
|
||||
if (status === "SUCCESS") return { completed: true, imageUrl };
|
||||
if (status === "FAILURE" || status === "CANCEL") {
|
||||
return { completed: false, error: `任务${status}: ${failReason || "未知原因"}` };
|
||||
}
|
||||
if (["NOT_START", "SUBMITTED", "IN_PROGRESS", "MODAL"].includes(status)) {
|
||||
return { completed: false };
|
||||
}
|
||||
return { completed: false, error: `未知状态: ${status}` };
|
||||
});
|
||||
} else if (manufacturer === "apimart") {
|
||||
// apimart 视频生成
|
||||
const apimartConfig = config as OpenAIVideoConfig;
|
||||
const apimartBaseURL = "https://api.apimart.ai";
|
||||
|
||||
// 上传图片到 apimart 图床
|
||||
let imageUrls: string[] = [];
|
||||
if (apimartConfig.imageBase64 && apimartConfig.imageBase64.length > 0) {
|
||||
for (const base64Image of apimartConfig.imageBase64) {
|
||||
// 如果已经是 URL，直接使用
|
||||
if (base64Image.startsWith("http")) {
|
||||
imageUrls.push(base64Image);
|
||||
continue;
|
||||
}
|
||||
|
||||
// 获取预签名 URL
|
||||
const presignRes = await axios.post(
|
||||
"https://apimart.ai/api/upload/presign",
|
||||
{ contentType: "image/jpeg", fileExtension: "jpeg", permanent: false },
|
||||
{ headers: { "Content-Type": "application/json" } },
|
||||
);
|
||||
|
||||
if (!presignRes.data.success || !presignRes.data.presignedUrl || !presignRes.data.cdnUrl) {
|
||||
throw new Error(`获取预签名 URL 失败: ${JSON.stringify(presignRes.data)}`);
|
||||
}
|
||||
|
||||
const { presignedUrl, cdnUrl } = presignRes.data;
|
||||
|
||||
// 移除 base64 前缀并转为 buffer
|
||||
const base64Data = base64Image.replace(/^data:image\/\w+;base64,/, "");
|
||||
const buffer = Buffer.from(base64Data, "base64");
|
||||
|
||||
// 上传图片到预签名 URL
|
||||
await axios.put(presignedUrl, buffer, {
|
||||
headers: { "Content-Type": "image/jpeg" },
|
||||
});
|
||||
|
||||
imageUrls.push(cdnUrl);
|
||||
}
|
||||
}
|
||||
|
||||
// 创建视频生成任务
|
||||
const requestBody: {
|
||||
model: string;
|
||||
prompt: string;
|
||||
duration: number;
|
||||
aspect_ratio: string;
|
||||
image_urls?: string[];
|
||||
} = {
|
||||
model: model || "sora-2",
|
||||
prompt: config.prompt,
|
||||
duration: apimartConfig.duration,
|
||||
aspect_ratio: apimartConfig.aspectRatio,
|
||||
};
|
||||
|
||||
if (imageUrls.length > 0) {
|
||||
requestBody.image_urls = imageUrls;
|
||||
}
|
||||
|
||||
const createRes = await axios.post(`${apimartBaseURL}/v1/videos/generations`, requestBody, {
|
||||
headers: {
|
||||
Authorization: `Bearer ${apiKey}`,
|
||||
"Content-Type": "application/json",
|
||||
},
|
||||
});
|
||||
|
||||
if (createRes.data.code !== 200 || !createRes.data.data?.[0]?.task_id) {
|
||||
const errorMsg = createRes.data.error?.message || JSON.stringify(createRes.data);
|
||||
throw new Error(`视频任务创建失败: ${errorMsg}`);
|
||||
}
|
||||
|
||||
const taskId = createRes.data.data[0].task_id;
|
||||
|
||||
// 轮询任务状态
|
||||
videoUrl = await pollTask(async () => {
|
||||
const res = await axios.get(`${apimartBaseURL}/v1/tasks/${taskId}`, {
|
||||
headers: { Authorization: `Bearer ${apiKey}` },
|
||||
params: { language: "en" },
|
||||
});
|
||||
|
||||
// 检查是否有错误
|
||||
if (res.data.error) {
|
||||
return {
|
||||
completed: false,
|
||||
error: `查询失败: ${res.data.error.message || JSON.stringify(res.data.error)}`,
|
||||
};
|
||||
}
|
||||
|
||||
if (res.data.code !== 200) {
|
||||
return { completed: false, error: `查询失败: ${JSON.stringify(res.data)}` };
|
||||
}
|
||||
|
||||
const { status, result } = res.data.data;
|
||||
|
||||
if (status === "completed") {
|
||||
// 获取视频 URL
|
||||
const videoUrlResult = result?.videos?.[0]?.url?.[0];
|
||||
return { completed: true, imageUrl: videoUrlResult };
|
||||
}
|
||||
|
||||
if (status === "failed" || status === "cancelled") {
|
||||
return { completed: false, error: `任务${status}` };
|
||||
}
|
||||
|
||||
// 其他状态（submitted, processing 等）继续轮询
|
||||
return { completed: false };
|
||||
});
|
||||
} else {
|
||||
throw new Error(`不支持的厂商: ${manufacturer}`);
|
||||
}
|
||||
return videoUrl;
|
||||
};
|
||||
export const generateVideo = async (config: VideoConfig, manufacturer: string) => {
|
||||
if (!config.imageBase64 || config.imageBase64.length <= 0) throw new Error("未传图片");
|
||||
const configList = await u.getConfig("video", manufacturer);
|
||||
console.log("%c Line:533 🥔 configList", "background:#ea7e5c", configList);
|
||||
if (!configList || configList.length === 0) {
|
||||
throw new Error("未找到任何视频配置");
|
||||
}
|
||||
let lastError: Error | null = null;
|
||||
for (const configItem of configList) {
|
||||
// 每个配置项重试1次，共2次尝试
|
||||
for (let attempt = 0; attempt < 2; attempt++) {
|
||||
try {
|
||||
const videoUrl = await generateVideoWithConfig(config, configItem);
|
||||
if (videoUrl) {
|
||||
const response = await axios.get(videoUrl, { responseType: "stream" });
|
||||
await u.oss.writeFile(config.savePath, response.data);
|
||||
return config.savePath;
|
||||
}
|
||||
return videoUrl;
|
||||
} catch (error: any) {
|
||||
lastError = error as Error;
|
||||
console.warn(`配置 ${configItem.model} 第 ${attempt + 1} 次尝试失败:`, error?.response?.data || error.message);
|
||||
// 如果是第一次尝试失败，继续重试
|
||||
if (attempt === 0) continue;
|
||||
// 第二次也失败了,跳到下一个配置项
|
||||
break;
|
||||
}
|
||||
}
|
||||
}
|
||||
// 所有配置都失败了
|
||||
throw new Error(`所有视频配置都失败了。最后一次错误: ${lastError?.message || "未知错误"}`);
|
||||
};
|
||||
@ -5,16 +5,11 @@ import axios from "axios";
|
||||
|
||||
import volcengine from "./owned/volcengine";
|
||||
import kling from "./owned/kling";
|
||||
import gemini from "./owned/gemini";
|
||||
import vidu from "./owned/vidu";
|
||||
import runninghub from "./owned/runninghub";
|
||||
import apimart from "./owned/apimart";
|
||||
import other from "./owned/other";
|
||||
interface AIConfig {
|
||||
model?: string;
|
||||
apiKey?: string;
|
||||
baseURL?: string;
|
||||
}
|
||||
import gemini from "./owned/gemini";
|
||||
|
||||
const urlToBase64 = async (url: string): Promise<string> => {
|
||||
const res = await axios.get(url, { responseType: "arraybuffer" });
|
||||
@ -29,8 +24,8 @@ const modelInstance = {
|
||||
kling: kling,
|
||||
vidu: vidu,
|
||||
runninghub: runninghub,
|
||||
apimart: apimart,
|
||||
other
|
||||
// apimart: apimart,
|
||||
other,
|
||||
} as const;
|
||||
|
||||
export default async (input: ImageConfig, config?: AIConfig) => {
|
||||
@ -38,11 +33,35 @@ export default async (input: ImageConfig, config?: AIConfig) => {
|
||||
const { model, apiKey, baseURL, manufacturer } = { ...sqlTextModelConfig, ...config };
|
||||
const manufacturerFn = modelInstance[manufacturer as keyof typeof modelInstance];
|
||||
if (!manufacturerFn) if (!manufacturerFn) throw new Error("不支持的图片厂商");
|
||||
// const owned = modelList.find((m) => m.model === model);
|
||||
// if (!owned) throw new Error("不支持的模型");
|
||||
const owned = modelList.find((m) => m.model === model);
|
||||
if (!owned) throw new Error("不支持的模型");
|
||||
|
||||
// 补充图片的 base64 内容类型字符串
|
||||
if (input.imageBase64 && input.imageBase64.length > 0) {
|
||||
input.imageBase64 = input.imageBase64.map((img) => {
|
||||
if (img.startsWith("data:image/")) {
|
||||
return img;
|
||||
}
|
||||
// 根据 base64 头部判断图片类型
|
||||
if (img.startsWith("/9j/")) {
|
||||
return `data:image/jpeg;base64,${img}`;
|
||||
}
|
||||
if (img.startsWith("iVBORw")) {
|
||||
return `data:image/png;base64,${img}`;
|
||||
}
|
||||
if (img.startsWith("R0lGOD")) {
|
||||
return `data:image/gif;base64,${img}`;
|
||||
}
|
||||
if (img.startsWith("UklGR")) {
|
||||
return `data:image/webp;base64,${img}`;
|
||||
}
|
||||
// 默认使用 png
|
||||
return `data:image/png;base64,${img}`;
|
||||
});
|
||||
}
|
||||
|
||||
let imageUrl = await manufacturerFn(input, { model, apiKey, baseURL });
|
||||
if (!input.resType) input.resType = "b64";
|
||||
if (input.resType === "b64" && imageUrl.startsWith("http")) imageUrl = await urlToBase64(imageUrl);
|
||||
return imageUrl;
|
||||
return input;
|
||||
};
|
||||
|
||||
@ -40,6 +40,12 @@ const modelList: Owned[] = [
|
||||
type: "ti2i",
|
||||
},
|
||||
//Vidu
|
||||
{
|
||||
manufacturer: "vidu",
|
||||
model: "viduq1",
|
||||
grid: false,
|
||||
type: "i2i",
|
||||
},
|
||||
{
|
||||
manufacturer: "vidu",
|
||||
model: "viduq2",
|
||||
|
||||
@ -24,7 +24,7 @@ export default async (input: ImageConfig, config: AIConfig): Promise<string> =>
|
||||
const res = await axios.get(`https://api.apimart.ai/v1/tasks/${taskId}`, { headers: { Authorization: apiKey }, params: { language: "en" } });
|
||||
if (res.data.code !== 200) return { completed: false, error: `查询失败: ${JSON.stringify(res.data)}` };
|
||||
const { status, result } = res.data.data;
|
||||
if (status === "completed") return { completed: true, imageUrl: result?.images?.[0]?.url?.[0] };
|
||||
if (status === "completed") return { completed: true, url: result?.images?.[0]?.url?.[0] };
|
||||
if (status === "failed" || status === "cancelled") return { completed: false, error: `任务${status}` };
|
||||
return { completed: false };
|
||||
});
|
||||
|
||||
@ -96,7 +96,7 @@ export default async (input: ImageConfig, config: AIConfig): Promise<string> =>
|
||||
}
|
||||
|
||||
if (task_status === "succeed") {
|
||||
return { completed: true, imageUrl: task_result?.images?.[0]?.url };
|
||||
return { completed: true, url: task_result?.images?.[0]?.url };
|
||||
}
|
||||
|
||||
return { completed: false };
|
||||
|
||||
@ -1,5 +1,4 @@
|
||||
import "../type";
|
||||
import { createGoogleGenerativeAI } from "@ai-sdk/google";
|
||||
import { generateImage, generateText } from "ai";
|
||||
import { createOpenAICompatible } from "@ai-sdk/openai-compatible";
|
||||
|
||||
|
||||
@ -85,7 +85,7 @@ export default async (input: ImageConfig, config: AIConfig): Promise<string> =>
|
||||
return pollTask(async () => {
|
||||
const res = await axios.post(`https://www.runninghub.cn/task/openapi/outputs`, { taskId, apiKey: apiKey });
|
||||
const { code, msg, data } = res.data;
|
||||
if (code === 0 && msg === "success") return { completed: true, imageUrl: data?.[0]?.fileUrl };
|
||||
if (code === 0 && msg === "success") return { completed: true, url: data?.[0]?.fileUrl };
|
||||
if (code === 804 || code === 813) return { completed: false };
|
||||
if (code === 805) return { completed: false, error: `任务失败: ${data?.[0]?.failedReason?.exception_message || "未知原因"}` };
|
||||
return { completed: false, error: `未知状态: code=${code}, msg=${msg}` };
|
||||
|
||||
@ -80,7 +80,7 @@ export default async (input: ImageConfig, config: AIConfig): Promise<string> =>
|
||||
}
|
||||
|
||||
if (state === "succeed") {
|
||||
return { completed: true, imageUrl: creations?.[0]?.url };
|
||||
return { completed: true, url: creations?.[0]?.url };
|
||||
}
|
||||
|
||||
return { completed: false };
|
||||
|
||||
@ -1,13 +1,76 @@
|
||||
import modelList from "./video/modelList";
|
||||
|
||||
interface ValidateResult {
|
||||
owned: (typeof modelList)[number];
|
||||
images: string[];
|
||||
hasStartEndType: boolean;
|
||||
hasTextType: boolean;
|
||||
}
|
||||
|
||||
/**
|
||||
* 校验视频生成配置与模型是否匹配
|
||||
* @param input 视频配置
|
||||
* @param config AI配置
|
||||
* @param customOwned 自定义模型配置（如果传入则跳过模型查找）
|
||||
*/
|
||||
export const validateVideoConfig = (input: VideoConfig, config: AIConfig, customOwned?: (typeof modelList)[number]): ValidateResult => {
|
||||
if (!config.model) throw new Error("缺少Model名称");
|
||||
const owned = customOwned ?? modelList.find((m) => m.model === config.model);
|
||||
if (!owned) throw new Error(`不支持的模型: ${config.model}`);
|
||||
const images = input.imageBase64 ?? [];
|
||||
// 校验图片数量与模型类型是否匹配
|
||||
const hasTextType = owned.type.includes("text");
|
||||
const hasSingleImageType = owned.type.includes("singleImage");
|
||||
const hasStartEndType = owned.type.some((t) => ["startEndRequired", "endFrameOptional", "startFrameOptional"].includes(t));
|
||||
const hasMultiImageType = owned.type.includes("multiImage");
|
||||
const hasReferenceType = owned.type.includes("reference");
|
||||
if (images.length === 0 && !hasTextType) {
|
||||
throw new Error(`模型 ${config.model} 不支持纯文本生成，需要提供图片`);
|
||||
}
|
||||
if (images.length === 1 && !hasSingleImageType && !hasStartEndType && !hasReferenceType) {
|
||||
throw new Error(`模型 ${config.model} 不支持单图模式`);
|
||||
}
|
||||
if (images.length === 2 && !hasStartEndType) {
|
||||
throw new Error(`模型 ${config.model} 不支持首尾帧模式`);
|
||||
}
|
||||
if (images.length > 2 && !hasMultiImageType) {
|
||||
throw new Error(`模型 ${config.model} 不支持多图模式`);
|
||||
}
|
||||
// 校验duration和resolution是否在支持范围内
|
||||
const validDurationResolution = owned.durationResolutionMap.some(
|
||||
(map) => map.duration.includes(input.duration) && map.resolution.includes(input.resolution as typeof map.resolution[number]),
|
||||
);
|
||||
if (!validDurationResolution) {
|
||||
const supportedDurations = [...new Set(owned.durationResolutionMap.flatMap((m) => m.duration))].sort((a, b) => a - b);
|
||||
const supportedResolutions = [...new Set(owned.durationResolutionMap.flatMap((m) => m.resolution))];
|
||||
throw new Error(
|
||||
`不支持的duration(${input.duration})或resolution(${input.resolution})组合。` +
|
||||
`支持的duration: ${supportedDurations.join(", ")}，支持的resolution: ${supportedResolutions.join(", ")}`,
|
||||
);
|
||||
}
|
||||
// 校验音频设置
|
||||
if (input.audio && !owned.audio) {
|
||||
throw new Error(`模型 ${config.model} 不支持生成音频`);
|
||||
}
|
||||
// 校验宽高比（仅文本生视频需要）
|
||||
if (hasTextType && images.length === 0 && owned.aspectRatio.length > 0) {
|
||||
if (!owned.aspectRatio.includes(input.aspectRatio as `${number}:${number}`)) {
|
||||
throw new Error(`模型 ${config.model} 不支持宽高比 ${input.aspectRatio}，支持的宽高比: ${owned.aspectRatio.join(", ")}`);
|
||||
}
|
||||
}
|
||||
return { owned, images, hasStartEndType, hasTextType };
|
||||
};
|
||||
|
||||
export const pollTask = async (
|
||||
queryFn: () => Promise<{ completed: boolean; imageUrl?: string; error?: string }>,
|
||||
queryFn: () => Promise<{ completed: boolean; url?: string; error?: string }>,
|
||||
maxAttempts = 500,
|
||||
interval = 2000,
|
||||
): Promise<string> => {
|
||||
for (let i = 0; i < maxAttempts; i++) {
|
||||
await new Promise((resolve) => setTimeout(resolve, interval));
|
||||
const { completed, imageUrl, error } = await queryFn();
|
||||
const { completed, url, error } = await queryFn();
|
||||
if (error) throw new Error(error);
|
||||
if (completed && imageUrl) return imageUrl;
|
||||
if (completed && url) return url;
|
||||
}
|
||||
throw new Error(`任务轮询超时，已尝试 ${maxAttempts} 次`);
|
||||
};
|
||||
};
|
||||
|
||||
63
src/utils/ai/video/index.ts
Normal file
63
src/utils/ai/video/index.ts
Normal file
@ -0,0 +1,63 @@
|
||||
import "./type";
|
||||
import u from "@/utils";
|
||||
import modelList from "./modelList";
|
||||
import axios from "axios";
|
||||
|
||||
import volcengine from "./owned/volcengine";
|
||||
import kling from "./owned/kling";
|
||||
import vidu from "./owned/vidu";
|
||||
import wan from "./owned/wan";
|
||||
import runninghub from "./owned/runninghub";
|
||||
import gemini from "./owned/gemini";
|
||||
import apimart from "./owned/apimart";
|
||||
|
||||
const modelInstance = {
|
||||
volcengine: volcengine,
|
||||
kling: kling,
|
||||
vidu: vidu,
|
||||
wan: wan,
|
||||
gemini: gemini,
|
||||
runninghub: runninghub,
|
||||
apimart: apimart,
|
||||
} as const;
|
||||
|
||||
export default async (input: VideoConfig, config?: AIConfig) => {
|
||||
const sqlTextModelConfig = await u.getConfig("video");
|
||||
const { model, apiKey, baseURL, manufacturer } = { ...sqlTextModelConfig, ...config };
|
||||
const manufacturerFn = modelInstance[manufacturer as keyof typeof modelInstance];
|
||||
if (!manufacturerFn) if (!manufacturerFn) throw new Error("不支持的视频厂商");
|
||||
const owned = modelList.find((m) => m.model === model);
|
||||
if (!owned) throw new Error("不支持的模型");
|
||||
|
||||
// 补充图片的 base64 内容类型字符串
|
||||
if (input.imageBase64 && input.imageBase64.length > 0) {
|
||||
input.imageBase64 = input.imageBase64.map((img) => {
|
||||
if (img.startsWith("data:image/")) {
|
||||
return img;
|
||||
}
|
||||
// 根据 base64 头部判断图片类型
|
||||
if (img.startsWith("/9j/")) {
|
||||
return `data:image/jpeg;base64,${img}`;
|
||||
}
|
||||
if (img.startsWith("iVBORw")) {
|
||||
return `data:image/png;base64,${img}`;
|
||||
}
|
||||
if (img.startsWith("R0lGOD")) {
|
||||
return `data:image/gif;base64,${img}`;
|
||||
}
|
||||
if (img.startsWith("UklGR")) {
|
||||
return `data:image/webp;base64,${img}`;
|
||||
}
|
||||
// 默认使用 png
|
||||
return `data:image/png;base64,${img}`;
|
||||
});
|
||||
}
|
||||
|
||||
let videoUrl = await manufacturerFn(input, { model, apiKey, baseURL });
|
||||
if (videoUrl) {
|
||||
const response = await axios.get(videoUrl, { responseType: "stream" });
|
||||
await u.oss.writeFile(input.savePath, response.data);
|
||||
return input.savePath;
|
||||
}
|
||||
return videoUrl;
|
||||
};
|
||||
489
src/utils/ai/video/modelList.ts
Normal file
489
src/utils/ai/video/modelList.ts
Normal file
@ -0,0 +1,489 @@
|
||||
type VideoGenerationType =
|
||||
| "singleImage" // 单图
|
||||
| "startEndRequired" // 首尾帧（两张都得有）
|
||||
| "endFrameOptional" // 首尾帧（尾帧可选）
|
||||
| "startFrameOptional" // 首尾帧（首帧可选）
|
||||
| "multiImage" // 多图模式
|
||||
| "reference" // 参考图模式
|
||||
| "text"; // 文本生视频
|
||||
|
||||
interface DurationResolutionMap {
|
||||
duration: number[];
|
||||
resolution: (`${number}p` | `${number}k`)[];
|
||||
}
|
||||
interface Owned {
|
||||
manufacturer: string;
|
||||
model: string;
|
||||
durationResolutionMap: DurationResolutionMap[];
|
||||
aspectRatio: `${number}:${number}`[];
|
||||
type: VideoGenerationType[];
|
||||
audio: boolean;
|
||||
}
|
||||
|
||||
const modelList: Owned[] = [
|
||||
// ================== 火山引擎/豆包系列 ==================
|
||||
// doubao-seedance-1-5-pro 文生视频/图生视频
|
||||
{
|
||||
manufacturer: "volcengine",
|
||||
model: "doubao-seedance-1-5-pro-251215",
|
||||
durationResolutionMap: [{ duration: [4, 5, 6, 7, 8, 9, 10, 11, 12], resolution: ["480p", "720p", "1080p"] }],
|
||||
aspectRatio: ["16:9", "4:3", "1:1", "3:4", "9:16", "21:9"],
|
||||
type: ["text", "endFrameOptional"],
|
||||
audio: true,
|
||||
},
|
||||
// doubao-seedance-1-0-pro 文生视频/图生视频
|
||||
{
|
||||
manufacturer: "volcengine",
|
||||
model: "doubao-seedance-1-0-pro-250528",
|
||||
durationResolutionMap: [{ duration: [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12], resolution: ["480p", "720p", "1080p"] }],
|
||||
aspectRatio: ["16:9", "4:3", "1:1", "3:4", "9:16", "21:9"],
|
||||
type: ["text", "endFrameOptional"],
|
||||
audio: false,
|
||||
},
|
||||
// doubao-seedance-1-0-pro-fast 文生视频/图生视频
|
||||
{
|
||||
manufacturer: "volcengine",
|
||||
model: "doubao-seedance-1-0-pro-fast-251015",
|
||||
durationResolutionMap: [{ duration: [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12], resolution: ["480p", "720p", "1080p"] }],
|
||||
aspectRatio: ["16:9", "4:3", "1:1", "3:4", "9:16", "21:9"],
|
||||
type: ["text", "singleImage"],
|
||||
audio: false,
|
||||
},
|
||||
// doubao-seedance-1-0-lite-i2v 图生视频（仅支持图片模式）
|
||||
{
|
||||
manufacturer: "volcengine",
|
||||
model: "doubao-seedance-1-0-lite-i2v-250428",
|
||||
durationResolutionMap: [{ duration: [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12], resolution: ["480p", "720p", "1080p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["endFrameOptional", "reference"],
|
||||
audio: false,
|
||||
},
|
||||
// doubao-seedance-1-0-lite-t2v 文生视频（仅支持文本模式）
|
||||
{
|
||||
manufacturer: "volcengine",
|
||||
model: "doubao-seedance-1-0-lite-t2v-250428",
|
||||
durationResolutionMap: [{ duration: [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12], resolution: ["480p", "720p", "1080p"] }],
|
||||
aspectRatio: ["16:9", "4:3", "1:1", "3:4", "9:16", "21:9"],
|
||||
type: ["text"],
|
||||
audio: false,
|
||||
},
|
||||
// ================== 可灵系列 ==================
|
||||
// kling-v1(STD) 文生视频
|
||||
{
|
||||
manufacturer: "kling",
|
||||
model: "kling-v1(STD)",
|
||||
durationResolutionMap: [{ duration: [5, 10], resolution: ["720p"] }],
|
||||
aspectRatio: ["16:9", "1:1", "9:16"],
|
||||
type: ["text"],
|
||||
audio: false,
|
||||
},
|
||||
// kling-v1(STD) 图生视频
|
||||
{
|
||||
manufacturer: "kling",
|
||||
model: "kling-v1(STD)",
|
||||
durationResolutionMap: [{ duration: [5, 10], resolution: ["720p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["startEndRequired"],
|
||||
audio: false,
|
||||
},
|
||||
// kling-v1(PRO) 文生视频
|
||||
{
|
||||
manufacturer: "kling",
|
||||
model: "kling-v1(PRO)",
|
||||
durationResolutionMap: [{ duration: [5, 10], resolution: ["1080p"] }],
|
||||
aspectRatio: ["16:9", "1:1", "9:16"],
|
||||
type: ["text"],
|
||||
audio: false,
|
||||
},
|
||||
// kling-v1(PRO) 图生视频
|
||||
{
|
||||
manufacturer: "kling",
|
||||
model: "kling-v1(PRO)",
|
||||
durationResolutionMap: [{ duration: [5, 10], resolution: ["1080p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["startEndRequired"],
|
||||
audio: false,
|
||||
},
|
||||
// kling-v1-6(PRO) 文生视频
|
||||
{
|
||||
manufacturer: "kling",
|
||||
model: "kling-v1-6(PRO)",
|
||||
durationResolutionMap: [{ duration: [5, 10], resolution: ["1080p"] }],
|
||||
aspectRatio: ["16:9", "1:1", "9:16"],
|
||||
type: ["text"],
|
||||
audio: false,
|
||||
},
|
||||
// kling-v1-6(PRO) 图生视频
|
||||
{
|
||||
manufacturer: "kling",
|
||||
model: "kling-v1-6(PRO)",
|
||||
durationResolutionMap: [{ duration: [5, 10], resolution: ["1080p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["startEndRequired"],
|
||||
audio: false,
|
||||
},
|
||||
// kling-v2-5-turbo(PRO) 文生视频
|
||||
{
|
||||
manufacturer: "kling",
|
||||
model: "kling-v2-5-turbo(PRO)",
|
||||
durationResolutionMap: [{ duration: [5, 10], resolution: ["1080p"] }],
|
||||
aspectRatio: ["16:9", "1:1", "9:16"],
|
||||
type: ["text"],
|
||||
audio: false,
|
||||
},
|
||||
// kling-v2-5-turbo(PRO) 图生视频
|
||||
{
|
||||
manufacturer: "kling",
|
||||
model: "kling-v2-5-turbo(PRO)",
|
||||
durationResolutionMap: [{ duration: [5, 10], resolution: ["1080p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["startEndRequired"],
|
||||
audio: false,
|
||||
},
|
||||
// kling-v2-6(PRO) 文生视频
|
||||
{
|
||||
manufacturer: "kling",
|
||||
model: "kling-v2-6(PRO)",
|
||||
durationResolutionMap: [{ duration: [5, 10], resolution: ["1080p"] }],
|
||||
aspectRatio: ["16:9", "1:1", "9:16"],
|
||||
type: ["text"],
|
||||
audio: false,
|
||||
},
|
||||
// kling-v2-6(PRO) 图生视频
|
||||
{
|
||||
manufacturer: "kling",
|
||||
model: "kling-v2-6(PRO)",
|
||||
durationResolutionMap: [{ duration: [5, 10], resolution: ["1080p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["startEndRequired"],
|
||||
audio: false,
|
||||
},
|
||||
// ================== ViduQ3系列 ==================
|
||||
// viduq3-pro 文生视频
|
||||
{
|
||||
manufacturer: "vidu",
|
||||
model: "viduq3-pro",
|
||||
durationResolutionMap: [{ duration: [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16], resolution: ["540p", "720p", "1080p"] }],
|
||||
aspectRatio: ["16:9", "9:16", "3:4", "4:3", "1:1"],
|
||||
type: ["text"],
|
||||
audio: true,
|
||||
},
|
||||
// viduq3-pro 图生视频
|
||||
{
|
||||
manufacturer: "vidu",
|
||||
model: "viduq3-pro",
|
||||
durationResolutionMap: [{ duration: [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16], resolution: ["540p", "720p", "1080p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["singleImage"],
|
||||
audio: true,
|
||||
},
|
||||
// viduq2-pro-fast 图生视频
|
||||
{
|
||||
manufacturer: "vidu",
|
||||
model: "viduq2-pro-fast",
|
||||
durationResolutionMap: [{ duration: [1, 2, 3, 4, 5, 6, 7, 8, 9, 10], resolution: ["720p", "1080p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["singleImage", "startEndRequired"],
|
||||
audio: false,
|
||||
},
|
||||
// viduq2-pro 文生视频
|
||||
{
|
||||
manufacturer: "vidu",
|
||||
model: "viduq2-pro",
|
||||
durationResolutionMap: [{ duration: [1, 2, 3, 4, 5, 6, 7, 8, 9, 10], resolution: ["540p", "720p", "1080p"] }],
|
||||
aspectRatio: ["16:9", "9:16", "3:4", "4:3", "1:1"],
|
||||
type: ["text"],
|
||||
audio: false,
|
||||
},
|
||||
// viduq2-pro 图生视频
|
||||
{
|
||||
manufacturer: "vidu",
|
||||
model: "viduq2-pro",
|
||||
durationResolutionMap: [{ duration: [1, 2, 3, 4, 5, 6, 7, 8, 9, 10], resolution: ["540p", "720p", "1080p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["singleImage", "reference", "startEndRequired"],
|
||||
audio: false,
|
||||
},
|
||||
// viduq2-turbo 文生视频
|
||||
{
|
||||
manufacturer: "vidu",
|
||||
model: "viduq2-turbo",
|
||||
durationResolutionMap: [{ duration: [1, 2, 3, 4, 5, 6, 7, 8, 9, 10], resolution: ["540p", "720p", "1080p"] }],
|
||||
aspectRatio: ["16:9", "9:16", "3:4", "4:3", "1:1"],
|
||||
type: ["text"],
|
||||
audio: false,
|
||||
},
|
||||
// viduq2-turbo 图生视频
|
||||
{
|
||||
manufacturer: "vidu",
|
||||
model: "viduq2-turbo",
|
||||
durationResolutionMap: [{ duration: [1, 2, 3, 4, 5, 6, 7, 8, 9, 10], resolution: ["540p", "720p", "1080p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["singleImage", "reference", "startEndRequired"],
|
||||
audio: false,
|
||||
},
|
||||
// viduq1 文生视频
|
||||
{
|
||||
manufacturer: "vidu",
|
||||
model: "viduq1",
|
||||
durationResolutionMap: [{ duration: [5], resolution: ["1080p"] }],
|
||||
aspectRatio: ["16:9", "9:16", "1:1"],
|
||||
type: ["text"],
|
||||
audio: false,
|
||||
},
|
||||
// viduq1 图生视频
|
||||
{
|
||||
manufacturer: "vidu",
|
||||
model: "viduq1",
|
||||
durationResolutionMap: [{ duration: [5], resolution: ["1080p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["singleImage", "reference", "startEndRequired"],
|
||||
audio: false,
|
||||
},
|
||||
// viduq1-classic 图生视频
|
||||
{
|
||||
manufacturer: "vidu",
|
||||
model: "viduq1-classic",
|
||||
durationResolutionMap: [{ duration: [5], resolution: ["1080p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["singleImage", "startEndRequired"],
|
||||
audio: false,
|
||||
},
|
||||
// vidu2.0 图生视频
|
||||
{
|
||||
manufacturer: "vidu",
|
||||
model: "vidu2.0",
|
||||
durationResolutionMap: [
|
||||
{ duration: [4], resolution: ["360p", "720p", "1080p"] },
|
||||
{ duration: [8], resolution: ["720p"] },
|
||||
],
|
||||
aspectRatio: [],
|
||||
type: ["singleImage", "reference", "startEndRequired"],
|
||||
audio: false,
|
||||
},
|
||||
// ================== 万象系列 ==================
|
||||
// wan2.6-t2v 文生视频（有声视频）
|
||||
{
|
||||
manufacturer: "wan",
|
||||
model: "wan2.6-t2v",
|
||||
durationResolutionMap: [{ duration: [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15], resolution: ["720p", "1080p"] }],
|
||||
aspectRatio: ["16:9", "9:16", "1:1", "4:3", "3:4"],
|
||||
type: ["text"],
|
||||
audio: true,
|
||||
},
|
||||
// wan2.5-t2v-preview 文生视频（有声视频）
|
||||
{
|
||||
manufacturer: "wan",
|
||||
model: "wan2.5-t2v-preview",
|
||||
durationResolutionMap: [{ duration: [5, 10], resolution: ["480p", "720p", "1080p"] }],
|
||||
aspectRatio: ["16:9", "9:16", "1:1", "4:3", "3:4"],
|
||||
type: ["text"],
|
||||
audio: true,
|
||||
},
|
||||
// wan2.2-t2v-plus 文生视频（无声视频）
|
||||
{
|
||||
manufacturer: "wan",
|
||||
model: "wan2.2-t2v-plus",
|
||||
durationResolutionMap: [{ duration: [5], resolution: ["480p", "1080p"] }],
|
||||
aspectRatio: ["16:9", "9:16", "1:1", "4:3", "3:4"],
|
||||
type: ["text"],
|
||||
audio: false,
|
||||
},
|
||||
// wanx2.1-t2v-turbo 文生视频（无声视频）
|
||||
{
|
||||
manufacturer: "wan",
|
||||
model: "wanx2.1-t2v-turbo",
|
||||
durationResolutionMap: [{ duration: [5], resolution: ["480p", "720p"] }],
|
||||
aspectRatio: ["16:9", "9:16", "1:1", "4:3", "3:4"],
|
||||
type: ["text"],
|
||||
audio: false,
|
||||
},
|
||||
// wanx2.1-t2v-plus 文生视频（无声视频）
|
||||
{
|
||||
manufacturer: "wan",
|
||||
model: "wanx2.1-t2v-plus",
|
||||
durationResolutionMap: [{ duration: [5], resolution: ["720p"] }],
|
||||
aspectRatio: ["16:9", "9:16", "1:1", "4:3", "3:4"],
|
||||
type: ["text"],
|
||||
audio: false,
|
||||
},
|
||||
// wan2.6-i2v-flash 图生视频（有声视频&无声视频）
|
||||
{
|
||||
manufacturer: "wan",
|
||||
model: "wan2.6-i2v-flash",
|
||||
durationResolutionMap: [{ duration: [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15], resolution: ["720p", "1080p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["singleImage"],
|
||||
audio: true,
|
||||
},
|
||||
// wan2.6-i2v 图生视频（有声视频）
|
||||
{
|
||||
manufacturer: "wan",
|
||||
model: "wan2.6-i2v",
|
||||
durationResolutionMap: [{ duration: [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15], resolution: ["720p", "1080p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["singleImage"],
|
||||
audio: true,
|
||||
},
|
||||
// wan2.5-i2v-preview 图生视频（有声视频）
|
||||
{
|
||||
manufacturer: "wan",
|
||||
model: "wan2.5-i2v-preview",
|
||||
durationResolutionMap: [{ duration: [5, 10], resolution: ["480p", "720p", "1080p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["singleImage"],
|
||||
audio: true,
|
||||
},
|
||||
// wan2.2-i2v-flash 图生视频（无声视频）
|
||||
{
|
||||
manufacturer: "wan",
|
||||
model: "wan2.2-i2v-flash",
|
||||
durationResolutionMap: [{ duration: [5], resolution: ["480p", "720p", "1080p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["singleImage"],
|
||||
audio: false,
|
||||
},
|
||||
// wan2.2-i2v-plus 图生视频（无声视频）
|
||||
{
|
||||
manufacturer: "wan",
|
||||
model: "wan2.2-i2v-plus",
|
||||
durationResolutionMap: [{ duration: [5], resolution: ["480p", "1080p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["singleImage"],
|
||||
audio: false,
|
||||
},
|
||||
// wanx2.1-i2v-plus 图生视频（无声视频）
|
||||
{
|
||||
manufacturer: "wan",
|
||||
model: "wanx2.1-i2v-plus",
|
||||
durationResolutionMap: [{ duration: [5], resolution: ["720p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["singleImage"],
|
||||
audio: false,
|
||||
},
|
||||
// wanx2.1-i2v-turbo 图生视频（无声视频）
|
||||
{
|
||||
manufacturer: "wan",
|
||||
model: "wanx2.1-i2v-turbo",
|
||||
durationResolutionMap: [{ duration: [3, 4, 5], resolution: ["480p", "720p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["singleImage"],
|
||||
audio: false,
|
||||
},
|
||||
// wan2.2-kf2v-flash 首尾帧生视频（无声视频）
|
||||
{
|
||||
manufacturer: "wan",
|
||||
model: "wan2.2-kf2v-flash",
|
||||
durationResolutionMap: [{ duration: [5], resolution: ["480p", "720p", "1080p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["startEndRequired"],
|
||||
audio: false,
|
||||
},
|
||||
// wanx2.1-kf2v-plus 首尾帧生视频（无声视频）
|
||||
{
|
||||
manufacturer: "wan",
|
||||
model: "wanx2.1-kf2v-plus",
|
||||
durationResolutionMap: [{ duration: [5], resolution: ["720p"] }],
|
||||
aspectRatio: [],
|
||||
type: ["startEndRequired"],
|
||||
audio: false,
|
||||
},
|
||||
// ================== Gemini Veo 系列 ==================
|
||||
// Veo 3.1 预览版（支持音频）
|
||||
{
|
||||
manufacturer: "gemini",
|
||||
model: "veo-3.1-generate-preview",
|
||||
durationResolutionMap: [
|
||||
{ duration: [4, 6], resolution: ["720p"] },
|
||||
{ duration: [8], resolution: ["720p", "1080p"] },
|
||||
],
|
||||
aspectRatio: ["16:9", "9:16"],
|
||||
type: ["text", "singleImage", "startEndRequired", "endFrameOptional", "reference"],
|
||||
audio: true,
|
||||
},
|
||||
// Veo 3.1 Fast 预览版（支持音频）
|
||||
{
|
||||
manufacturer: "gemini",
|
||||
model: "veo-3.1-fast-generate-preview",
|
||||
durationResolutionMap: [
|
||||
{ duration: [4, 6], resolution: ["720p"] },
|
||||
{ duration: [8], resolution: ["720p", "1080p"] },
|
||||
],
|
||||
aspectRatio: ["16:9", "9:16"],
|
||||
type: ["text", "singleImage", "startEndRequired", "endFrameOptional", "reference"],
|
||||
audio: true,
|
||||
},
|
||||
// Veo 3 稳定版（支持音频）
|
||||
{
|
||||
manufacturer: "gemini",
|
||||
model: "veo-3.0-generate-preview",
|
||||
durationResolutionMap: [
|
||||
{ duration: [4, 6], resolution: ["720p"] },
|
||||
{ duration: [8], resolution: ["720p", "1080p"] },
|
||||
],
|
||||
aspectRatio: ["16:9", "9:16"],
|
||||
type: ["text", "singleImage"],
|
||||
audio: true,
|
||||
},
|
||||
// Veo 3 Fast 稳定版（支持音频）
|
||||
{
|
||||
manufacturer: "gemini",
|
||||
model: "veo-3.0-fast-generate-preview",
|
||||
durationResolutionMap: [
|
||||
{ duration: [4, 6], resolution: ["720p"] },
|
||||
{ duration: [8], resolution: ["720p", "1080p"] },
|
||||
],
|
||||
aspectRatio: ["16:9", "9:16"],
|
||||
type: ["text", "singleImage"],
|
||||
audio: true,
|
||||
},
|
||||
// Veo 2 稳定版（无音频）
|
||||
{
|
||||
manufacturer: "gemini",
|
||||
model: "veo-2.0-generate-001",
|
||||
durationResolutionMap: [{ duration: [5, 6, 7, 8], resolution: ["720p"] }],
|
||||
aspectRatio: ["16:9", "9:16"],
|
||||
type: ["text", "singleImage"],
|
||||
audio: false,
|
||||
},
|
||||
// ================== RunningHub 系列 ==================
|
||||
// sora
|
||||
{
|
||||
manufacturer: "runninghub",
|
||||
model: "sora-2",
|
||||
durationResolutionMap: [{ duration: [10, 15], resolution: [] }],
|
||||
aspectRatio: ["16:9", "9:16"],
|
||||
type: ["singleImage", "text"],
|
||||
audio: false,
|
||||
},
|
||||
// sora 2
|
||||
{
|
||||
manufacturer: "runninghub",
|
||||
model: "sora-2-pro",
|
||||
durationResolutionMap: [{ duration: [15, 25], resolution: [] }],
|
||||
aspectRatio: ["16:9", "9:16"],
|
||||
type: ["singleImage", "text"],
|
||||
audio: false,
|
||||
},
|
||||
// ================== Apimart 系列 ==================
|
||||
// sora
|
||||
{
|
||||
manufacturer: "apimart",
|
||||
model: "sora-2",
|
||||
durationResolutionMap: [{ duration: [10, 15], resolution: [] }],
|
||||
aspectRatio: ["16:9", "9:16"],
|
||||
type: ["singleImage", "text"],
|
||||
audio: false,
|
||||
},
|
||||
// sora 2
|
||||
{
|
||||
manufacturer: "apimart",
|
||||
model: "sora-2-pro",
|
||||
durationResolutionMap: [{ duration: [15, 25], resolution: [] }],
|
||||
aspectRatio: ["16:9", "9:16"],
|
||||
type: ["singleImage", "text"],
|
||||
audio: false,
|
||||
},
|
||||
];
|
||||
|
||||
export default modelList;
|
||||
115
src/utils/ai/video/owned/apimart.ts
Normal file
115
src/utils/ai/video/owned/apimart.ts
Normal file
@ -0,0 +1,115 @@
|
||||
import "../type";
|
||||
import axios from "axios";
|
||||
import { pollTask } from "@/utils/ai/utils";
|
||||
import modelList from "../modelList";
|
||||
|
||||
// 上传图片到 apimart 图床
|
||||
async function uploadImageToApimart(base64Image: string): Promise<string> {
|
||||
if (base64Image.startsWith("http")) {
|
||||
return base64Image;
|
||||
}
|
||||
|
||||
const presignRes = await axios.post(
|
||||
"https://apimart.ai/api/upload/presign",
|
||||
{ contentType: "image/jpeg", fileExtension: "jpeg", permanent: false },
|
||||
{ headers: { "Content-Type": "application/json" } },
|
||||
);
|
||||
|
||||
if (!presignRes.data.success || !presignRes.data.presignedUrl || !presignRes.data.cdnUrl) {
|
||||
throw new Error(`获取预签名 URL 失败: ${JSON.stringify(presignRes.data)}`);
|
||||
}
|
||||
|
||||
const { presignedUrl, cdnUrl } = presignRes.data;
|
||||
|
||||
const base64Data = base64Image.replace(/^data:image\/\w+;base64,/, "");
|
||||
const buffer = Buffer.from(base64Data, "base64");
|
||||
|
||||
await axios.put(presignedUrl, buffer, {
|
||||
headers: { "Content-Type": "image/jpeg" },
|
||||
});
|
||||
|
||||
return cdnUrl;
|
||||
}
|
||||
|
||||
export default async (input: VideoConfig, config: AIConfig) => {
|
||||
if (!config.model) throw new Error("缺少 Model 名称");
|
||||
if (!config.apiKey) throw new Error("缺少 API Key");
|
||||
|
||||
const owned = modelList.find((m) => m.model === config.model);
|
||||
if (!owned) throw new Error(`未找到模型: ${config.model}`);
|
||||
|
||||
// 默认 baseURL 配置
|
||||
const defaultBaseUrl = "https://api.apimart.ai/v1/videos/generations|https://api.apimart.ai/v1/tasks/{taskId}";
|
||||
const [generateUrl, queryUrl] = (config.baseURL || defaultBaseUrl).split("|");
|
||||
|
||||
const authorization = `Bearer ${config.apiKey}`;
|
||||
|
||||
// 上传图片到图床
|
||||
let imageUrls: string[] = [];
|
||||
if (input.imageBase64 && input.imageBase64.length > 0) {
|
||||
for (const base64Image of input.imageBase64) {
|
||||
const imageUrl = await uploadImageToApimart(base64Image);
|
||||
imageUrls.push(imageUrl);
|
||||
}
|
||||
}
|
||||
|
||||
// 构建请求体
|
||||
const requestBody: Record<string, unknown> = {
|
||||
model: config.model,
|
||||
prompt: input.prompt,
|
||||
duration: input.duration,
|
||||
aspect_ratio: input.aspectRatio,
|
||||
};
|
||||
|
||||
if (imageUrls.length > 0) {
|
||||
requestBody.image_urls = imageUrls;
|
||||
}
|
||||
|
||||
// 创建任务
|
||||
const createRes = await axios.post(generateUrl, requestBody, {
|
||||
headers: {
|
||||
Authorization: authorization,
|
||||
"Content-Type": "application/json",
|
||||
},
|
||||
});
|
||||
|
||||
if (createRes.data.code !== 200 || !createRes.data.data?.[0]?.task_id) {
|
||||
throw new Error(`创建任务失败: ${JSON.stringify(createRes.data)}`);
|
||||
}
|
||||
|
||||
const taskId = createRes.data.data[0].task_id;
|
||||
const actualQueryUrl = queryUrl.replace("{taskId}", taskId);
|
||||
|
||||
// 轮询任务状态
|
||||
return await pollTask(async () => {
|
||||
const queryRes = await axios.get(actualQueryUrl, {
|
||||
headers: { Authorization: authorization },
|
||||
});
|
||||
|
||||
const { code, data } = queryRes.data;
|
||||
|
||||
if (code !== 200 || !data) {
|
||||
return { completed: false, error: `查询失败: ${JSON.stringify(queryRes.data)}` };
|
||||
}
|
||||
|
||||
const { status, result, error } = data;
|
||||
|
||||
switch (status) {
|
||||
case "completed":
|
||||
const videoUrl = result?.videos?.[0]?.url?.[0];
|
||||
if (!videoUrl) {
|
||||
return { completed: false, error: "未获取到视频 URL" };
|
||||
}
|
||||
return { completed: true, url: videoUrl };
|
||||
case "failed":
|
||||
return { completed: false, error: error?.message || "任务失败" };
|
||||
case "cancelled":
|
||||
return { completed: false, error: "任务已取消" };
|
||||
case "pending":
|
||||
case "processing":
|
||||
return { completed: false };
|
||||
default:
|
||||
return { completed: false, error: `未知状态: ${status}` };
|
||||
}
|
||||
});
|
||||
};
|
||||
68
src/utils/ai/video/owned/gemini.ts
Normal file
68
src/utils/ai/video/owned/gemini.ts
Normal file
@ -0,0 +1,68 @@
|
||||
import "../type";
|
||||
import fs from "fs";
|
||||
import path from "path";
|
||||
import axios from "axios";
|
||||
import { pollTask, validateVideoConfig } from "@/utils/ai/utils";
|
||||
|
||||
const buildInlineImage = (data: string) => ({ inlineData: { mimeType: "image/png", data } });
|
||||
|
||||
export default async (input: VideoConfig, config: AIConfig) => {
|
||||
if (!config.model) throw new Error("缺少Model名称");
|
||||
if (!config.apiKey) throw new Error("缺少API Key");
|
||||
|
||||
const { owned, images, hasStartEndType } = validateVideoConfig(input, config);
|
||||
|
||||
const defaultBaseUrl = [
|
||||
"https://generativelanguage.googleapis.com/v1beta/models/{model}:predictLongRunning",
|
||||
"https://generativelanguage.googleapis.com/v1beta/{name}",
|
||||
].join("|");
|
||||
|
||||
const [submitUrl, queryUrl] = (config.baseURL || defaultBaseUrl).split("|");
|
||||
const headers = { "x-goog-api-key": config.apiKey };
|
||||
|
||||
const instance: Record<string, any> = { prompt: input.prompt };
|
||||
const parameters: Record<string, any> = {
|
||||
aspectRatio: input.aspectRatio,
|
||||
durationSeconds: String(input.duration),
|
||||
...(input.resolution !== "720p" && { resolution: input.resolution }),
|
||||
};
|
||||
|
||||
// 根据图片数量和模型能力决定图片用法
|
||||
const len = images.length;
|
||||
const hasRef = owned.type.includes("reference");
|
||||
const hasSingle = owned.type.includes("singleImage");
|
||||
|
||||
if (len === 2 && hasStartEndType) {
|
||||
instance.image = buildInlineImage(images[0]);
|
||||
parameters.lastFrame = buildInlineImage(images[1]);
|
||||
} else if (len === 1 && (hasSingle || hasStartEndType)) {
|
||||
instance.image = buildInlineImage(images[0]);
|
||||
} else if (len >= 1 && len <= 3 && hasRef) {
|
||||
parameters.referenceImages = images.map((img) => ({ image: buildInlineImage(img), referenceType: "asset" }));
|
||||
}
|
||||
|
||||
const { data } = await axios.post(
|
||||
submitUrl.replace("{model}", config.model),
|
||||
{ instances: [instance], parameters },
|
||||
{ headers: { ...headers, "Content-Type": "application/json" } },
|
||||
);
|
||||
|
||||
if (!data.name) throw new Error("未获取到操作名称");
|
||||
|
||||
return pollTask(async () => {
|
||||
const { data: status } = await axios.get(queryUrl.replace("{name}", data.name), { headers });
|
||||
const { done, response, error } = status;
|
||||
|
||||
if (!done) return { completed: false };
|
||||
if (error) return { completed: false, error: `任务失败: ${error.message || JSON.stringify(error)}` };
|
||||
|
||||
const videoUri = response?.generateVideoResponse?.generatedSamples?.[0]?.video?.uri;
|
||||
if (!videoUri) return { completed: false, error: "未获取到视频下载地址" };
|
||||
|
||||
const videoRes = await axios.get(videoUri, { headers, responseType: "arraybuffer", maxRedirects: 5 });
|
||||
const savePath = input.savePath.endsWith(".mp4") ? input.savePath : path.join(input.savePath, `gemini_${Date.now()}.mp4`);
|
||||
fs.writeFileSync(savePath, Buffer.from(videoRes.data));
|
||||
|
||||
return { completed: true, url: savePath };
|
||||
});
|
||||
};
|
||||
90
src/utils/ai/video/owned/kling.ts
Normal file
90
src/utils/ai/video/owned/kling.ts
Normal file
@ -0,0 +1,90 @@
|
||||
import "../type";
|
||||
import axios from "axios";
|
||||
import { pollTask, validateVideoConfig } from "@/utils/ai/utils";
|
||||
|
||||
export default async (input: VideoConfig, config: AIConfig) => {
|
||||
if (!config.apiKey) throw new Error("缺少API Key");
|
||||
if (!config.baseURL) throw new Error("缺少baseURL配置");
|
||||
|
||||
const { images } = validateVideoConfig(input, config);
|
||||
|
||||
// 解析URL配置：图生视频|文生视频|查询地址
|
||||
const defaultBaseUrl =
|
||||
"https://api-beijing.klingai.com/v1/videos/image2video|https://api-beijing.klingai.com/v1/videos/text2video|https://api-beijing.klingai.com/v1/videos/text2video/{taskId}";
|
||||
const [image2videoUrl, text2videoUrl, queryUrl] = (config.baseURL || defaultBaseUrl).split("|");
|
||||
|
||||
const headers = {
|
||||
Authorization: `Bearer ${config.apiKey}`,
|
||||
"Content-Type": "application/json",
|
||||
};
|
||||
|
||||
// 解析模型名称和模式，例如 "kling-v2-6(PRO)" => modelName: "kling-v2-6", mode: "pro"
|
||||
const modelMatch = config.model!.match(/^(.+)\((STD|PRO)\)$/i);
|
||||
const modelName = modelMatch ? modelMatch[1] : config.model;
|
||||
const mode = modelMatch ? (modelMatch[2].toLowerCase() as "std" | "pro") : "std";
|
||||
|
||||
// 判断是图生视频还是文生视频
|
||||
const hasImage = images.length > 0;
|
||||
const createUrl = hasImage ? image2videoUrl : text2videoUrl;
|
||||
|
||||
// 去除图片的内容类型前缀（kling要求纯base64）
|
||||
const stripDataUrl = (str: string) => str.replace(/^data:image\/[^;]+;base64,/, "");
|
||||
|
||||
// 构建请求体
|
||||
const body: Record<string, unknown> = {
|
||||
model_name: modelName,
|
||||
mode,
|
||||
duration: String(input.duration),
|
||||
prompt: input.prompt,
|
||||
aspect_ratio: input.aspectRatio,
|
||||
};
|
||||
|
||||
if (hasImage) {
|
||||
// 图生视频：首帧和尾帧
|
||||
body.image = stripDataUrl(images[0]);
|
||||
if (images.length > 1) {
|
||||
body.image_tail = stripDataUrl(images[1]);
|
||||
}
|
||||
}
|
||||
|
||||
// 创建任务
|
||||
const createResponse = await axios.post(createUrl, body, { headers });
|
||||
const createData = createResponse.data;
|
||||
if (createData.code !== 0) {
|
||||
throw new Error(`创建任务失败: ${createData.message || "未知错误"}`);
|
||||
}
|
||||
|
||||
const taskId = createData.data?.task_id;
|
||||
if (!taskId) {
|
||||
throw new Error("创建任务失败: 未返回任务ID");
|
||||
}
|
||||
|
||||
// 轮询任务状态
|
||||
return await pollTask(async () => {
|
||||
const queryResponse = await axios.get(`${queryUrl.replace("{taskId}", taskId)}`, { headers });
|
||||
const queryData = queryResponse.data;
|
||||
if (queryData.code !== 0) {
|
||||
return { completed: false, error: `查询失败: ${queryData.message || "未知错误"}` };
|
||||
}
|
||||
|
||||
const task = queryData.data;
|
||||
const taskStatus = task?.task_status;
|
||||
|
||||
switch (taskStatus) {
|
||||
case "succeed": {
|
||||
const videoUrl = task?.task_result?.videos?.[0]?.url;
|
||||
if (!videoUrl) {
|
||||
return { completed: false, error: "任务成功但未返回视频URL" };
|
||||
}
|
||||
return { completed: true, url: videoUrl };
|
||||
}
|
||||
case "failed":
|
||||
return { completed: false, error: `任务失败: ${task?.task_status_msg || "未知原因"}` };
|
||||
case "submitted":
|
||||
case "processing":
|
||||
return { completed: false };
|
||||
default:
|
||||
return { completed: false, error: `未知状态: ${taskStatus}` };
|
||||
}
|
||||
});
|
||||
};
|
||||
91
src/utils/ai/video/owned/runninghub.ts
Normal file
91
src/utils/ai/video/owned/runninghub.ts
Normal file
@ -0,0 +1,91 @@
|
||||
import "../type";
|
||||
import axios from "axios";
|
||||
import sharp from "sharp";
|
||||
import FormData from "form-data";
|
||||
import { pollTask, validateVideoConfig } from "@/utils/ai/utils";
|
||||
|
||||
export default async (input: VideoConfig, config: AIConfig) => {
|
||||
if (!config.apiKey) throw new Error("缺少API Key");
|
||||
|
||||
const { owned, images, hasTextType } = validateVideoConfig(input, config);
|
||||
|
||||
const defaultBaseUrl = [
|
||||
"https://www.runninghub.cn/openapi/v2/rhart-video-s/image-to-video",
|
||||
"https://www.runninghub.cn/openapi/v2/rhart-video-s/image-to-video-pro",
|
||||
"https://www.runninghub.cn/openapi/v2/rhart-video-s/text-to-video",
|
||||
"https://www.runninghub.cn/openapi/v2/rhart-video-s/text-to-video-pro",
|
||||
"https://www.runninghub.cn/openapi/v2/rhart-video-s/{taskId}",
|
||||
"https://www.runninghub.cn/openapi/v2/media/upload/binary",
|
||||
].join("|");
|
||||
|
||||
const [image2videoUrl, image2videoProUrl, text2videoUrl, text2videoProUrl, queryUrl, uploadUrl] = (config.baseURL || defaultBaseUrl).split("|");
|
||||
|
||||
const isPro = owned.model === "sora-2-pro";
|
||||
const authorization = `Bearer ${config.apiKey}`;
|
||||
|
||||
// 上传 base64 图片
|
||||
const uploadImage = async (base64Image: string): Promise<string> => {
|
||||
const base64Data = base64Image.replace(/^data:image\/\w+;base64,/, "");
|
||||
let buffer: Buffer = Buffer.from(base64Data, "base64");
|
||||
const MAX_SIZE = 5 * 1024 * 1024;
|
||||
|
||||
if (buffer.length > MAX_SIZE) {
|
||||
for (let quality = 90; buffer.length > MAX_SIZE && quality > 10; quality -= 10) {
|
||||
buffer = await sharp(buffer).jpeg({ quality, mozjpeg: true }).toBuffer();
|
||||
}
|
||||
if (buffer.length > MAX_SIZE) {
|
||||
const { width = 1920, height = 1080 } = await sharp(buffer).metadata();
|
||||
const scale = Math.sqrt(MAX_SIZE / buffer.length);
|
||||
buffer = await sharp(buffer)
|
||||
.resize({ width: Math.floor(width * scale), height: Math.floor(height * scale), fit: "inside" })
|
||||
.jpeg({ quality: 80, mozjpeg: true })
|
||||
.toBuffer();
|
||||
}
|
||||
}
|
||||
|
||||
const formData = new FormData();
|
||||
formData.append("file", buffer, { filename: "image.jpg", contentType: "image/jpeg" });
|
||||
|
||||
const { data } = await axios.post(uploadUrl, formData, {
|
||||
headers: { Authorization: authorization },
|
||||
});
|
||||
|
||||
if (data.code !== 0 || !data.data?.download_url) {
|
||||
throw new Error(`图片上传失败: ${JSON.stringify(data)}`);
|
||||
}
|
||||
return data.data.download_url;
|
||||
};
|
||||
|
||||
// 提交任务
|
||||
const submitTask = async (url: string, body: Record<string, unknown>) => {
|
||||
const { data } = await axios.post(url, body, {
|
||||
headers: { "Content-Type": "application/json", Authorization: authorization },
|
||||
});
|
||||
if (data.status === "FAILED") throw new Error(`任务提交失败: ${data.errorMessage || "未知错误"}`);
|
||||
return { taskId: data.taskId, status: data.status, url: data.results?.[0]?.url };
|
||||
};
|
||||
|
||||
const isTextToVideo = images.length === 0 && hasTextType;
|
||||
const submitUrl = isTextToVideo ? (isPro ? text2videoProUrl : text2videoUrl) : isPro ? image2videoProUrl : image2videoUrl;
|
||||
|
||||
const requestBody: Record<string, unknown> = {
|
||||
prompt: input.prompt,
|
||||
duration: String(input.duration),
|
||||
aspectRatio: input.aspectRatio,
|
||||
...(isTextToVideo ? {} : { imageUrl: await uploadImage(images[0]) }),
|
||||
};
|
||||
|
||||
const { taskId } = await submitTask(submitUrl, requestBody);
|
||||
|
||||
return await pollTask(async () => {
|
||||
const { data } = await axios.get(queryUrl.replace("{taskId}", taskId), {
|
||||
headers: { Authorization: authorization },
|
||||
});
|
||||
if (data.status === "SUCCESS") {
|
||||
return data.results?.length ? { completed: true, url: data.results[0].url } : { completed: false, error: "任务成功但未返回视频链接" };
|
||||
}
|
||||
if (data.status === "FAILED") return { completed: false, error: `任务失败: ${data.errorMessage || "未知错误"}` };
|
||||
if (data.status === "QUEUED" || data.status === "RUNNING") return { completed: false };
|
||||
return { completed: false, error: `未知状态: ${data.status}` };
|
||||
});
|
||||
};
|
||||
132
src/utils/ai/video/owned/vidu.ts
Normal file
132
src/utils/ai/video/owned/vidu.ts
Normal file
@ -0,0 +1,132 @@
|
||||
import "../type";
|
||||
import axios from "axios";
|
||||
import { pollTask, validateVideoConfig } from "@/utils/ai/utils";
|
||||
import modelList from "../modelList";
|
||||
|
||||
export default async (input: VideoConfig, config: AIConfig) => {
|
||||
if (!config.model) throw new Error("缺少Model名称");
|
||||
if (!config.apiKey) throw new Error("缺少API Key");
|
||||
if (!input.prompt && (!input.imageBase64 || input.imageBase64.length === 0)) {
|
||||
throw new Error("至少需要提供prompt或图片");
|
||||
}
|
||||
|
||||
const defaultBaseUrl = ["https://api.vidu.cn/ent/v2/text2video", "https://api.vidu.cn/ent/v2/img2video", "https://api.vidu.cn/ent/v2/tasks"].join(
|
||||
"|",
|
||||
);
|
||||
|
||||
const [text2videoUrl, image2videoUrl, queryUrl] = (config.baseURL || defaultBaseUrl).split("|");
|
||||
|
||||
const authorization = `Token ${config.apiKey}`;
|
||||
const hasImages = input.imageBase64 && input.imageBase64.length > 0;
|
||||
|
||||
// 根据是否有图片，查找匹配的模型配置
|
||||
const customOwned = modelList.find((m) => {
|
||||
if (m.manufacturer !== "vidu") return false;
|
||||
if (m.model !== config.model) return false;
|
||||
if (hasImages) {
|
||||
return m.type.some((t) => t !== "text");
|
||||
} else {
|
||||
return m.type.includes("text");
|
||||
}
|
||||
});
|
||||
|
||||
if (!customOwned) {
|
||||
throw new Error(`未找到匹配的模型配置: ${config.model}`);
|
||||
}
|
||||
|
||||
// 使用统一校验函数
|
||||
const { owned, images } = validateVideoConfig(input, config, customOwned);
|
||||
|
||||
// 判断生成类型
|
||||
const genType: "text" | "image" = images.length === 0 ? "text" : "image";
|
||||
|
||||
// 校验宽高比（仅文生视频需要）
|
||||
if (genType === "text" && owned.aspectRatio.length > 0 && !owned.aspectRatio.includes(input.aspectRatio as `${number}:${number}`)) {
|
||||
throw new Error(`模型 ${owned.model} 不支持宽高比 ${input.aspectRatio}，支持的宽高比：${owned.aspectRatio.join("、")}`);
|
||||
}
|
||||
|
||||
// 创建任务
|
||||
let taskId: string;
|
||||
|
||||
if (genType === "text") {
|
||||
// 文生视频
|
||||
const requestBody: Record<string, unknown> = {
|
||||
model: owned.model,
|
||||
prompt: input.prompt,
|
||||
duration: input.duration,
|
||||
resolution: input.resolution,
|
||||
aspect_ratio: input.aspectRatio,
|
||||
};
|
||||
if (owned.audio && input.audio !== undefined) {
|
||||
requestBody.audio = input.audio;
|
||||
}
|
||||
|
||||
const response = await axios.post(text2videoUrl, requestBody, {
|
||||
headers: {
|
||||
"Content-Type": "application/json",
|
||||
Authorization: authorization,
|
||||
},
|
||||
});
|
||||
taskId = response.data.task_id;
|
||||
} else {
|
||||
// 图生视频
|
||||
const requestBody: Record<string, unknown> = {
|
||||
model: owned.model,
|
||||
images: images,
|
||||
duration: input.duration,
|
||||
resolution: input.resolution,
|
||||
};
|
||||
if (input.prompt) {
|
||||
requestBody.prompt = input.prompt;
|
||||
}
|
||||
if (owned.audio && input.audio !== undefined) {
|
||||
requestBody.audio = input.audio;
|
||||
}
|
||||
|
||||
const response = await axios.post(image2videoUrl, requestBody, {
|
||||
headers: {
|
||||
"Content-Type": "application/json",
|
||||
Authorization: authorization,
|
||||
},
|
||||
});
|
||||
taskId = response.data.task_id;
|
||||
}
|
||||
|
||||
// 轮询任务状态
|
||||
return await pollTask(async () => {
|
||||
const response = await axios.get(queryUrl, {
|
||||
headers: {
|
||||
"Content-Type": "application/json",
|
||||
Authorization: authorization,
|
||||
},
|
||||
params: {
|
||||
task_ids: [taskId],
|
||||
},
|
||||
});
|
||||
|
||||
const tasks = response.data.tasks;
|
||||
if (!tasks || tasks.length === 0) {
|
||||
return { completed: false, error: "任务不存在" };
|
||||
}
|
||||
|
||||
const task = tasks[0];
|
||||
|
||||
switch (task.state) {
|
||||
case "success": {
|
||||
const creation = task.creations?.[0];
|
||||
return {
|
||||
completed: true,
|
||||
url: creation?.url,
|
||||
};
|
||||
}
|
||||
case "failed":
|
||||
return { completed: false, error: "任务生成失败" };
|
||||
case "created":
|
||||
case "queueing":
|
||||
case "processing":
|
||||
return { completed: false };
|
||||
default:
|
||||
return { completed: false, error: `未知状态: ${task.state}` };
|
||||
}
|
||||
});
|
||||
};
|
||||
@ -1,56 +1,74 @@
|
||||
import "../type";
|
||||
import axios from "axios";
|
||||
import { pollTask } from "@/utils/ai/utils";
|
||||
import { pollTask, validateVideoConfig } from "@/utils/ai/utils";
|
||||
|
||||
interface DoubaoVideoConfig {
|
||||
prompt: string;
|
||||
savePath: string;
|
||||
imageBase64?: string[]; // 单张参考图片 base64
|
||||
duration: 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12; // 支持 2~12 秒
|
||||
aspectRatio: "16:9" | "9:16" | "1:1" | "4:3" | "3:4" | "21:9" | "adaptive";
|
||||
audio?: boolean;
|
||||
}
|
||||
|
||||
export default async (input: ImageConfig, config: AIConfig) => {
|
||||
console.log("%c Line:5 🍓 input", "background:#7f2b82", input);
|
||||
console.log("%c Line:5 🍎 config", "background:#93c0a4", config);
|
||||
if (!config.model) throw new Error("缺少Model名称");
|
||||
export default async (input: VideoConfig, config: AIConfig) => {
|
||||
if (!config.apiKey) throw new Error("缺少API Key");
|
||||
|
||||
const key = "Bearer " + config.apiKey.replaceAll("Bearer ", "").trim();
|
||||
const { owned, images, hasStartEndType } = validateVideoConfig(input, config);
|
||||
|
||||
const doubaoConfig = config as DoubaoVideoConfig;
|
||||
const createRes = await axios.post(
|
||||
config.baseURL ?? "https://ark.cn-beijing.volces.com/api/v3/contents/generations/tasks",
|
||||
{
|
||||
model: "doubao-seedance-1-5-pro-251215",
|
||||
content: [
|
||||
{ type: "text", text: input.prompt },
|
||||
...(doubaoConfig.imageBase64
|
||||
? doubaoConfig.imageBase64.map((base64, i) => ({
|
||||
type: "image_url",
|
||||
image_url: { url: base64 },
|
||||
role: i === 0 ? "first_frame" : "last_frame",
|
||||
}))
|
||||
: []),
|
||||
],
|
||||
generate_audio: doubaoConfig.audio ?? false,
|
||||
duration: doubaoConfig.duration,
|
||||
resolution: doubaoConfig.aspectRatio,
|
||||
watermark: false,
|
||||
const authorization = "Bearer " + config.apiKey.replace(/^Bearer\s*/i, "").trim();
|
||||
const baseUrl = config.baseURL ?? "https://ark.cn-beijing.volces.com/api/v3/contents/generations/tasks";
|
||||
|
||||
// 判断是否为首尾帧模式（需要两张图且类型支持首尾帧）
|
||||
const isStartEndMode = images.length === 2 && hasStartEndType;
|
||||
|
||||
// 构建图片内容
|
||||
const imageContent = images.map((base64, index) => {
|
||||
const item: Record<string, any> = {
|
||||
type: "image_url",
|
||||
image_url: { url: base64 },
|
||||
};
|
||||
if (isStartEndMode) {
|
||||
item.role = index === 0 ? "first_frame" : "last_frame";
|
||||
}
|
||||
return item;
|
||||
});
|
||||
|
||||
// 构建请求体
|
||||
const requestBody: Record<string, any> = {
|
||||
model: config.model,
|
||||
content: [{ type: "text", text: input.prompt }, ...imageContent],
|
||||
duration: input.duration,
|
||||
resolution: input.resolution,
|
||||
watermark: false,
|
||||
};
|
||||
|
||||
// 仅当模型支持音频时才添加 generate_audio 字段
|
||||
if (owned.audio) {
|
||||
requestBody.generate_audio = input.audio ?? false;
|
||||
}
|
||||
// 创建视频生成任务
|
||||
const createResponse = await axios.post(baseUrl, requestBody, {
|
||||
headers: {
|
||||
"Content-Type": "application/json",
|
||||
Authorization: authorization,
|
||||
},
|
||||
{ headers: { "Content-Type": "application/json", Authorization: key } },
|
||||
);
|
||||
const taskId = createRes.data.id;
|
||||
});
|
||||
|
||||
const taskId = createResponse.data.id;
|
||||
if (!taskId) throw new Error("视频任务创建失败");
|
||||
|
||||
// 轮询任务状态
|
||||
return await pollTask(async () => {
|
||||
const res = await axios.get(`${config.baseURL ?? "https://ark.cn-beijing.volces.com/api/v3/contents/generations/tasks"}/${taskId}`, {
|
||||
headers: { Authorization: key },
|
||||
});
|
||||
const { status, content } = res.data;
|
||||
if (status === "succeeded") return { completed: true, imageUrl: content?.video_url };
|
||||
if (["failed", "cancelled", "expired"].includes(status)) return { completed: false, error: `任务${status}` };
|
||||
if (["queued", "running"].includes(status)) return { completed: false };
|
||||
return { completed: false, error: `未知状态: ${status}` };
|
||||
const { status, content } = (
|
||||
await axios.get(`${baseUrl}/${taskId}`, {
|
||||
headers: { Authorization: authorization },
|
||||
})
|
||||
).data;
|
||||
|
||||
switch (status) {
|
||||
case "succeeded":
|
||||
return { completed: true, url: content?.video_url };
|
||||
case "failed":
|
||||
case "cancelled":
|
||||
case "expired":
|
||||
return { completed: false, error: `任务${status}` };
|
||||
case "queued":
|
||||
case "running":
|
||||
return { completed: false };
|
||||
default:
|
||||
return { completed: false, error: `未知状态: ${status}` };
|
||||
}
|
||||
});
|
||||
};
|
||||
|
||||
168
src/utils/ai/video/owned/wan.ts
Normal file
168
src/utils/ai/video/owned/wan.ts
Normal file
@ -0,0 +1,168 @@
|
||||
import "../type";
|
||||
import axios from "axios";
|
||||
import { pollTask, validateVideoConfig } from "@/utils/ai/utils";
|
||||
|
||||
// 根据分辨率档位和宽高比计算具体尺寸
|
||||
const getSizeFromConfig = (resolution: string, aspectRatio: string): string => {
|
||||
const sizeMap: Record<string, Record<string, string>> = {
|
||||
"480p": {
|
||||
"16:9": "832*480",
|
||||
"9:16": "480*832",
|
||||
"1:1": "624*624",
|
||||
},
|
||||
"720p": {
|
||||
"16:9": "1280*720",
|
||||
"9:16": "720*1280",
|
||||
"1:1": "960*960",
|
||||
"4:3": "1088*832",
|
||||
"3:4": "832*1088",
|
||||
},
|
||||
"1080p": {
|
||||
"16:9": "1920*1080",
|
||||
"9:16": "1080*1920",
|
||||
"1:1": "1440*1440",
|
||||
"4:3": "1632*1248",
|
||||
"3:4": "1248*1632",
|
||||
},
|
||||
};
|
||||
|
||||
const resolutionKey = resolution.toLowerCase();
|
||||
const size = sizeMap[resolutionKey]?.[aspectRatio];
|
||||
|
||||
if (!size) {
|
||||
throw new Error(`不支持的分辨率(${resolution})和宽高比(${aspectRatio})组合`);
|
||||
}
|
||||
|
||||
return size;
|
||||
};
|
||||
|
||||
export default async (input: VideoConfig, config: AIConfig) => {
|
||||
if (!config.apiKey) throw new Error("缺少API Key");
|
||||
|
||||
const { owned, images, hasStartEndType, hasTextType } = validateVideoConfig(input, config);
|
||||
|
||||
const defaultBaseUrl = [
|
||||
"https://dashscope.aliyuncs.com/api/v1/services/aigc/video-generation/video-synthesis",
|
||||
"https://dashscope.aliyuncs.com/api/v1/services/aigc/image2video/video-synthesis",
|
||||
"https://dashscope.aliyuncs.com/api/v1/tasks/{taskId}",
|
||||
].join("|");
|
||||
|
||||
const [i2vUrl, kf2vUrl, queryUrl] = (config.baseURL || defaultBaseUrl).split("|");
|
||||
|
||||
const types = owned.type;
|
||||
const authorization = `Bearer ${config.apiKey}`;
|
||||
|
||||
// 确定端点和构建请求体
|
||||
let submitUrl: string;
|
||||
let body: Record<string, any>;
|
||||
|
||||
if (hasTextType && images.length === 0) {
|
||||
// 文本生视频
|
||||
submitUrl = i2vUrl;
|
||||
body = {
|
||||
model: config.model,
|
||||
input: {
|
||||
prompt: input.prompt,
|
||||
},
|
||||
parameters: {
|
||||
size: getSizeFromConfig(input.resolution, input.aspectRatio),
|
||||
duration: input.duration,
|
||||
},
|
||||
};
|
||||
} else if (types.includes("singleImage")) {
|
||||
// 图生视频
|
||||
submitUrl = i2vUrl;
|
||||
body = {
|
||||
model: config.model,
|
||||
input: {
|
||||
prompt: input.prompt,
|
||||
img_url: images[0],
|
||||
},
|
||||
parameters: {
|
||||
resolution: input.resolution.toUpperCase(),
|
||||
duration: input.duration,
|
||||
},
|
||||
};
|
||||
// audio参数仅部分模型支持
|
||||
if (owned.audio && input.audio !== undefined) {
|
||||
body.parameters.audio = input.audio;
|
||||
}
|
||||
} else if (hasStartEndType) {
|
||||
// 首尾帧
|
||||
submitUrl = kf2vUrl;
|
||||
const inputObj: Record<string, any> = {
|
||||
prompt: input.prompt,
|
||||
first_frame_url: images[0],
|
||||
};
|
||||
// 尾帧处理
|
||||
if (types.includes("startEndRequired")) {
|
||||
inputObj.last_frame_url = images[1];
|
||||
} else if ((types.includes("endFrameOptional") || types.includes("startFrameOptional")) && images.length >= 2) {
|
||||
inputObj.last_frame_url = images[1];
|
||||
}
|
||||
body = {
|
||||
model: config.model,
|
||||
input: inputObj,
|
||||
parameters: {
|
||||
resolution: input.resolution.toUpperCase(),
|
||||
duration: input.duration,
|
||||
},
|
||||
};
|
||||
} else {
|
||||
throw new Error(`不支持的视频生成类型: ${types.join(", ")}`);
|
||||
}
|
||||
|
||||
// 提交任务
|
||||
const submitResponse = await axios.post(submitUrl, body, {
|
||||
headers: {
|
||||
"Content-Type": "application/json",
|
||||
Authorization: authorization,
|
||||
"X-DashScope-Async": "enable",
|
||||
},
|
||||
});
|
||||
|
||||
const submitData = submitResponse.data;
|
||||
if (submitData.code) {
|
||||
throw new Error(`任务提交失败: [${submitData.code}] ${submitData.message}`);
|
||||
}
|
||||
|
||||
const taskId = submitData.output?.task_id;
|
||||
if (!taskId) {
|
||||
throw new Error("任务提交失败: 未返回task_id");
|
||||
}
|
||||
|
||||
// 轮询任务状态
|
||||
return await pollTask(async () => {
|
||||
const response = await axios.get(queryUrl.replace("{taskId}", taskId), {
|
||||
headers: { Authorization: authorization },
|
||||
});
|
||||
|
||||
const data = response.data;
|
||||
|
||||
// 顶层错误
|
||||
if (data.code) {
|
||||
return { completed: false, error: `[${data.code}] ${data.message}` };
|
||||
}
|
||||
|
||||
const taskStatus = data.output?.task_status;
|
||||
|
||||
switch (taskStatus) {
|
||||
case "SUCCEEDED":
|
||||
return { completed: true, url: data.output?.video_url };
|
||||
case "FAILED":
|
||||
return {
|
||||
completed: false,
|
||||
error: `任务失败: [${data.output?.code || "UNKNOWN"}] ${data.output?.message || "未知错误"}`,
|
||||
};
|
||||
case "CANCELED":
|
||||
return { completed: false, error: "任务已取消" };
|
||||
case "UNKNOWN":
|
||||
return { completed: false, error: "任务不存在或状态未知" };
|
||||
case "PENDING":
|
||||
case "RUNNING":
|
||||
return { completed: false };
|
||||
default:
|
||||
return { completed: false, error: `未知状态: ${taskStatus}` };
|
||||
}
|
||||
});
|
||||
};
|
||||
15
src/utils/ai/video/type.ts
Normal file
15
src/utils/ai/video/type.ts
Normal file
@ -0,0 +1,15 @@
|
||||
interface VideoConfig {
|
||||
duration: 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12;
|
||||
resolution: "480p" | "720p" | "1080p" | "2K" | "4K";
|
||||
aspectRatio: "16:9" | "9:16";
|
||||
prompt: string;
|
||||
savePath: string;
|
||||
imageBase64?: string[];
|
||||
audio?: boolean;
|
||||
}
|
||||
|
||||
interface AIConfig {
|
||||
model?: string;
|
||||
apiKey?: string;
|
||||
baseURL?: string;
|
||||
}
|
||||
Loading…
x
Reference in New Issue
Block a user