Video Generation

Video generation is asynchronous. POST /v1/videos/generations submits the task and returns an id; poll GET /v1/videos/{video_id} until status is "completed", then download the file via GET /v1/videos/{video_id} /content.

Parameters

Endpoint: POST /v1/videos/generations

Parameter	Type	Required	Description
`model`	`string`	Yes	Model ID, e.g. `wan-2.7`, `hailuo-2.3`, `veo-2.0-generate-001`, `gen4_turbo`, `seedance-01-lite`, `kling-v2-master`, `seedance-2.0`, `veo-3.1-generate`.
`prompt`	`string`	Yes	Text description of the desired video.
`input_reference`	`string \| file`	No	Reference image URL or uploaded file for image-to-video mode.
`seconds`	`integer`	No	Duration in seconds.
`size`	`string`	No	Legacy resolution parameter, e.g. `1920x1080`. Prefer `extra_body.resolution` with `extra_body.aspect_ratio`.
`extra_body`	`object`	No	Provider-specific extension parameters (see below).
`user`	`string`	No	End-user identifier, passed through as-is.

Passing Reference Assets

Goal	Field to pass
Single image-to-video request with one reference image	`input_reference`
Multiple reference images, or explicit multimodal reference mode	`extra_body.reference_image_urls`
Use reference videos to guide generation	`extra_body.reference_video_urls`
Use reference audio to guide generation	`extra_body.reference_audio_urls`
Specify first frame or start/end frames	`extra_body.first_frame_url` / `extra_body.last_frame_url`

input_reference and reference_image_urls have different semantics even when they contain the same image. input_reference is the primary single reference image for standard image-to-video requests. reference_image_urls is a multimodal reference list that the model may use for subject, style, composition, or consistency. You can pass one image in reference_image_urls, but only do this when you explicitly want multimodal reference behavior. If extra_body.reference_image_urls is set explicitly, input_reference will not override it.

extra_body Common Fields

Field	Type	Description
`resolution`	`string`	Resolution: `480p`, `720p`, `1080p`, or `4k`.
`aspect_ratio`	`string`	Aspect ratio: `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, or `21:9`.
`audio_url`	`string`	External audio URL.
`generate_audio`	`boolean`	Whether to automatically generate audio for the video.
`video_url`	`string`	Reference video URL for reference-to-video mode with Seedance.
`first_frame_url`	`string`	First-frame URL for first-frame or start/end-frame generation.
`first_image`	`string`	Alternative first-frame field accepted by compatible providers.
`last_frame_url`	`string`	Last-frame URL for start/end-frame generation.
`last_image`	`string`	Alternative last-frame field accepted by compatible providers.
`reference_image_urls`	`string[]`	Multimodal reference image URLs.
`reference_video_urls`	`string[]`	Multimodal reference video URLs.
`reference_audio_urls`	`string[]`	Multimodal reference audio URLs.
`motion_ref_url`	`string`	Motion action or movement reference video URL.
`motion_ref_urls`	`string[]`	Motion action or movement reference video URLs.
`motion_mode`	`string`	Motion generation mode. Kling supports `std` / `pro`; Seedance ignores this field.
`motion_strength`	`number`	Motion transfer strength. Valid ranges depend on provider capabilities.
`character_orientation`	`string`	Kling motion character orientation: `image` / `video`.
`keep_original_sound`	`boolean`	Whether Kling motion keeps the original video audio.
`camera_control`	`object`	Kling motion camera control.
`element_list`	`list`	Kling motion element list.
`background_tolerance`	`number`	Kling motion background tolerance.
`enhance_prompt`	`boolean`	Prompt enhancement for Veo. Defaults to `true`.
`prompt_optimization`	`boolean`	Prompt optimization for Pixwith. Defaults to `true`.
`subject_reference`	`list`	Subject references for Flora.
`watermark`	`boolean`	Watermark option for Seedance. Defaults to `false`.

When using motion features, input_reference must also be provided.

Model Capability Fields

Call /model_group/info before submitting a request to check whether a model supports a specific video capability.

Field	Type	Description
`supports_video_multimodal_reference`	`boolean`	Whether the model supports `reference_image_urls`, `reference_video_urls`, and `reference_audio_urls`.
`supports_video_start_end_frame`	`boolean`	Whether the model supports start/end-frame generation.

How It Works

Video generation follows a three-step flow:

Step 1 — Submit the task (POST /v1/videos/generations)

The request returns immediately with an id and status: "processing".

Step 2 — Poll for the result (GET /v1/videos/{video_id})

Repeat the poll request every few seconds until status is "completed".

Step 3 — Download the video (GET /v1/videos/{video_id}/content)

Fetch the binary video file once the task is complete.

Status	Meaning
`processing`	Video is being generated
`completed`	Done — ready to download
`failed`	Generation failed — see `error` field

All Language Examples

curl -X POST https://silkdock.ai/v1/videos/generations \-H "Authorization: Bearer $SILKDOCK_API_KEY" \-H "Content-Type: application/json" \-d '{  "model": "seedance-2.0",  "prompt": "Ocean waves crashing on rocks, sunlight refracting into rainbows",  "seconds": 10,  "extra_body": {    "resolution": "1080p",    "aspect_ratio": "16:9"  }}'

curl -X POST https://silkdock.ai/v1/videos/generations ^-H "Authorization: Bearer %SILKDOCK_API_KEY%" ^-H "Content-Type: application/json" ^-d "{"model":"seedance-2.0","prompt":"Ocean waves crashing on rocks","seconds":10,"extra_body":{"resolution":"1080p","aspect_ratio":"16:9"}}"

http POST https://silkdock.ai/v1/videos/generations \Authorization:"Bearer $SILKDOCK_API_KEY" \model=seedance-2.0 \prompt="Ocean waves crashing on rocks, sunlight refracting into rainbows" \seconds:=10 \extra_body:='{"resolution":"1080p","aspect_ratio":"16:9"}'

wget -q -O - https://silkdock.ai/v1/videos/generations \--method=POST \--header="Authorization: Bearer $SILKDOCK_API_KEY" \--header="Content-Type: application/json" \--body-data='{  "model": "seedance-2.0",  "prompt": "Ocean waves crashing on rocks, sunlight refracting into rainbows",  "seconds": 10,  "extra_body": {    "resolution": "1080p",    "aspect_ratio": "16:9"  }}'

$headers = @{  "Authorization" = "Bearer $env:SILKDOCK_API_KEY"  "Content-Type"  = "application/json"}$body = @{  model      = "seedance-2.0"  prompt     = "Ocean waves crashing on rocks, sunlight refracting into rainbows"  seconds    = 10  extra_body = @{      resolution   = "1080p"      aspect_ratio = "16:9"  }} | ConvertTo-Json -Depth 5$res = Invoke-RestMethod -Method POST `-Uri "https://silkdock.ai/v1/videos/generations" `-Headers $headers `-Body $bodyWrite-Output $res.id

const res = await fetch("https://silkdock.ai/v1/videos/generations", {method: "POST",headers: {  "Authorization": `Bearer ${process.env.SILKDOCK_API_KEY}`,  "Content-Type": "application/json",},body: JSON.stringify({  model: "seedance-2.0",  prompt: "Ocean waves crashing on rocks, sunlight refracting into rainbows",  seconds: 10,  extra_body: {    resolution: "1080p",    aspect_ratio: "16:9",  },}),});const data = await res.json();console.log(data.id);

import axios from "axios";const { data } = await axios.post("https://silkdock.ai/v1/videos/generations",{model: "seedance-2.0",prompt: "Ocean waves crashing on rocks, sunlight refracting into rainbows",seconds: 10,extra_body: {resolution: "1080p",aspect_ratio: "16:9",},},{headers: {Authorization: `Bearer ${process.env.SILKDOCK_API_KEY}`,"Content-Type": "application/json",},});console.log(data.id);

$.ajax({url: "https://silkdock.ai/v1/videos/generations",method: "POST",contentType: "application/json",headers: {Authorization: `Bearer ${SILKDOCK_API_KEY}`,},data: JSON.stringify({model: "seedance-2.0",prompt: "Ocean waves crashing on rocks, sunlight refracting into rainbows",seconds: 10,extra_body: {resolution: "1080p",aspect_ratio: "16:9",},}),success(res) {console.log(res.id);},});

const xhr = new XMLHttpRequest();xhr.open("POST", "https://silkdock.ai/v1/videos/generations");xhr.setRequestHeader("Authorization", `Bearer ${process.env.SILKDOCK_API_KEY}`);xhr.setRequestHeader("Content-Type", "application/json");xhr.onload = () => {const data = JSON.parse(xhr.responseText);console.log(data.id);};xhr.send(JSON.stringify({model: "seedance-2.0",prompt: "Ocean waves crashing on rocks, sunlight refracting into rainbows",seconds: 10,extra_body: {resolution: "1080p",aspect_ratio: "16:9",},}));

const request = require("request");request.post({url: "https://silkdock.ai/v1/videos/generations",headers: {Authorization: `Bearer ${process.env.SILKDOCK_API_KEY}`,"Content-Type": "application/json",},json: {model: "seedance-2.0",prompt: "Ocean waves crashing on rocks, sunlight refracting into rainbows",seconds: 10,extra_body: {resolution: "1080p",aspect_ratio: "16:9",},},},(err, res, body) => {console.log(body.id);});

const unirest = require("unirest");unirest.post("https://silkdock.ai/v1/videos/generations").headers({Authorization: `Bearer ${process.env.SILKDOCK_API_KEY}`,"Content-Type": "application/json",}).send({model: "seedance-2.0",prompt: "Ocean waves crashing on rocks, sunlight refracting into rainbows",seconds: 10,extra_body: {resolution: "1080p",aspect_ratio: "16:9",},}).then((res) => {console.log(res.body.id);});

import OpenAI from "openai";const client = new OpenAI({apiKey: process.env.SILKDOCK_API_KEY,baseURL: "https://silkdock.ai/v1",});const video = await client.videos.create({model: "seedance-2.0",prompt: "Ocean waves crashing on rocks, sunlight refracting into rainbows",seconds: 10,extra_body: {resolution: "1080p",aspect_ratio: "16:9",},});let result = video;while (result.status === "processing") {await new Promise((resolve) => setTimeout(resolve, 5000));result = await client.videos.retrieve(video.id);}console.log("Done:", result.status);

const res = await fetch("https://silkdock.ai/v1/videos/generations", {method: "POST",headers: {"Authorization": `Bearer ${process.env.SILKDOCK_API_KEY}`,"Content-Type": "application/json",},body: JSON.stringify({model: "seedance-2.0",prompt: "Ocean waves crashing on rocks, sunlight refracting into rainbows",seconds: 10,extra_body: {resolution: "1080p",aspect_ratio: "16:9",},}),});const { id } = await res.json() as { id: string };console.log(id);

from openai import OpenAIimport osimport timeclient = OpenAI(api_key=os.getenv("SILKDOCK_API_KEY"),base_url="https://silkdock.ai/v1",)video = client.videos.create(model="seedance-2.0",prompt="Ocean waves crashing on rocks, sunlight refracting into rainbows",seconds=10,extra_body={"resolution": "1080p","aspect_ratio": "16:9",},)while True:result = client.videos.retrieve(video.id)if result.status == "completed":breakif result.status == "failed":raise RuntimeError(result.error)time.sleep(5)content = client.videos.retrieve_content(video.id)with open("out.mp4", "wb") as f:f.write(content.read())

import requests, osres = requests.post("https://silkdock.ai/v1/videos/generations",headers={"Authorization": f"Bearer {os.getenv('SILKDOCK_API_KEY')}"},json={"model": "seedance-2.0","prompt": "Ocean waves crashing on rocks, sunlight refracting into rainbows","seconds": 10,"extra_body": {"resolution": "1080p","aspect_ratio": "16:9"},},)print(res.json()["id"])

#include <stdio.h>#include <curl/curl.h>int main(void) {CURL *curl = curl_easy_init();if (!curl) return 1;  struct curl_slist *headers = NULL;  headers = curl_slist_append(headers, "Content-Type: application/json");  headers = curl_slist_append(headers,      "Authorization: Bearer " SILKDOCK_API_KEY);  const char *body =      "{"model":"seedance-2.0","      ""prompt":"Ocean waves crashing on rocks","      ""seconds":10,"      ""extra_body":{"resolution":"1080p","aspect_ratio":"16:9"}}";  curl_easy_setopt(curl, CURLOPT_URL, "https://silkdock.ai/v1/videos/generations");  curl_easy_setopt(curl, CURLOPT_HTTPHEADER, headers);  curl_easy_setopt(curl, CURLOPT_POSTFIELDS, body);  curl_easy_perform(curl);  curl_slist_free_all(headers);  curl_easy_cleanup(curl);  return 0;}/_ compile: gcc main.c -lcurl -o main _/

#import <Foundation/Foundation.h>NSURLSession *session = [NSURLSession sharedSession];NSURL *url = [NSURL URLWithString:@"https://silkdock.ai/v1/videos/generations"];NSMutableURLRequest *request = [NSMutableURLRequest requestWithURL:url];[request setHTTPMethod:@"POST"];[request setValue:@"application/json" forHTTPHeaderField:@"Content-Type"];[request setValue:[@"Bearer " stringByAppendingString:[NSProcessInfo.processInfo.environment objectForKey:@"SILKDOCK_API_KEY"]]forHTTPHeaderField:@"Authorization"];NSDictionary *payload = @{@"model": @"seedance-2.0",@"prompt": @"Ocean waves crashing on rocks, sunlight refracting into rainbows",@"seconds": @10,@"extra_body": @{@"resolution": @"1080p",@"aspect_ratio": @"16:9"}};[request setHTTPBody:[NSJSONSerialization dataWithJSONObject:payload options:0 error:nil]];NSURLSessionDataTask *task = [session dataTaskWithRequest:requestcompletionHandler:^(NSData *data, NSURLResponse *response, NSError *error) {NSDictionary *json = [NSJSONSerialization JSONObjectWithData:data options:0 error:nil];NSLog(@"Video ID: %@", json[@"id"]);}];[task resume];

import java.net.http.*;import java.net.URI;var req = HttpRequest.newBuilder().uri(URI.create("https://silkdock.ai/v1/videos/generations")).header("Authorization", "Bearer " + System.getenv("SILKDOCK_API_KEY")).header("Content-Type", "application/json").POST(HttpRequest.BodyPublishers.ofString("""{"model":"seedance-2.0","prompt":"Ocean waves crashing on rocks, sunlight refracting into rainbows","seconds":10,"extra_body":{"resolution":"1080p","aspect_ratio":"16:9"}}""")).build();System.out.println(HttpClient.newHttpClient().send(req, HttpResponse.BodyHandlers.ofString()).body());

import okhttp3.*;OkHttpClient client = new OkHttpClient();RequestBody body = RequestBody.create("{"model":"seedance-2.0","prompt":"Ocean waves crashing on rocks"," +""seconds":10,"extra_body":{"resolution":"1080p","aspect_ratio":"16:9"}}",MediaType.get("application/json"));Request request = new Request.Builder().url("https://silkdock.ai/v1/videos/generations").addHeader("Authorization", "Bearer " + System.getenv("SILKDOCK_API_KEY")).post(body).build();try (Response response = client.newCall(request).execute()) {System.out.println(response.body().string());}

import kong.unirest.Unirest;var res = Unirest.post("https://silkdock.ai/v1/videos/generations").header("Authorization", "Bearer " + System.getenv("SILKDOCK_API_KEY")).header("Content-Type", "application/json").body("{"model":"seedance-2.0","prompt":"Ocean waves crashing on rocks"," +""seconds":10,"extra_body":{"resolution":"1080p","aspect_ratio":"16:9"}}").asString();System.out.println(res.getBody());

package mainimport (  "bytes"  "encoding/json"  "fmt"  "io"  "net/http"  "os")func main() {body, _ := json.Marshal(map[string]any{"model": "seedance-2.0","prompt": "Ocean waves crashing on rocks, sunlight refracting into rainbows","seconds": 10,"extra_body": map[string]any{"resolution": "1080p","aspect_ratio": "16:9",},})req, _ := http.NewRequest("POST", "https://silkdock.ai/v1/videos/generations", bytes.NewReader(body))req.Header.Set("Authorization", "Bearer "+os.Getenv("SILKDOCK*API_KEY"))req.Header.Set("Content-Type", "application/json")resp, * := http.DefaultClient.Do(req)defer resp.Body.Close()data, _ := io.ReadAll(resp.Body)fmt.Println(string(data))}

<?php$ch = curl_init("https://silkdock.ai/v1/videos/generations");curl_setopt_array($ch, [  CURLOPT_POST           => true,  CURLOPT_RETURNTRANSFER => true,  CURLOPT_HTTPHEADER     => [      "Authorization: Bearer " . getenv("SILKDOCK_API_KEY"),      "Content-Type: application/json",  ],  CURLOPT_POSTFIELDS => json_encode([      "model"   => "seedance-2.0",      "prompt"  => "Ocean waves crashing on rocks, sunlight refracting into rainbows",      "seconds" => 10,      "extra_body" => [          "resolution"   => "1080p",          "aspect_ratio" => "16:9",      ],  ]),]);$res = json_decode(curl_exec($ch), true);echo $res["id"];

<?phprequire_once "HTTP/Request2.php";$request = new HTTP_Request2(  "https://silkdock.ai/v1/videos/generations",  HTTP_Request2::METHOD_POST);$request->setHeader(["Authorization" => "Bearer " . getenv("SILKDOCK_API_KEY"),"Content-Type" => "application/json",]);$request->setBody(json_encode([  "model"   => "seedance-2.0",  "prompt"  => "Ocean waves crashing on rocks, sunlight refracting into rainbows",  "seconds" => 10,  "extra_body" => [      "resolution"   => "1080p",      "aspect_ratio" => "16:9",  ],]));$res = json_decode($request->send()->getBody(), true);echo $res["id"];

<?phpuse GuzzleHttpClient;$client = new Client();$res = $client->post("https://silkdock.ai/v1/videos/generations", [  "headers" => [      "Authorization" => "Bearer " . getenv("SILKDOCK_API_KEY"),      "Content-Type"  => "application/json",  ],  "json" => [      "model"   => "seedance-2.0",      "prompt"  => "Ocean waves crashing on rocks, sunlight refracting into rainbows",      "seconds" => 10,      "extra_body" => [          "resolution"   => "1080p",          "aspect_ratio" => "16:9",      ],  ],]);$body = json_decode($res->getBody(), true);echo $body["id"];

<?php$client = new httpClient;$request = new httpClientRequest(  "POST",  "https://silkdock.ai/v1/videos/generations");$request->setHeaders(["Authorization" => "Bearer " . getenv("SILKDOCK_API_KEY"),"Content-Type" => "application/json",]);$body = new httpMessageBody;$body->append(json_encode(["model" => "seedance-2.0","prompt" => "Ocean waves crashing on rocks, sunlight refracting into rainbows","seconds" => 10,"extra_body" => ["resolution" => "1080p","aspect_ratio" => "16:9",],]));$request->setBody($body);$client->enqueue($request)->send();$res = json_decode($client->getResponse()->getBody(), true);echo $res["id"];

import Foundationvar req = URLRequest(url: URL(string: "https://silkdock.ai/v1/videos/generations")!)req.httpMethod = "POST"req.setValue("Bearer \(ProcessInfo.processInfo.environment["SILKDOCK_API_KEY"]!)",forHTTPHeaderField: "Authorization")req.setValue("application/json", forHTTPHeaderField: "Content-Type")req.httpBody = try! JSONSerialization.data(withJSONObject: ["model": "seedance-2.0","prompt": "Ocean waves crashing on rocks, sunlight refracting into rainbows","seconds": 10,"extra*body": ["resolution": "1080p","aspect_ratio": "16:9"],])let (data, *) = try! await URLSession.shared.data(for: req)print(String(data: data, encoding: .utf8)!)

using System.Net.Http;using System.Net.Http.Json;var client = new HttpClient();client.DefaultRequestHeaders.Add("Authorization",$"Bearer {Environment.GetEnvironmentVariable("SILKDOCK_API_KEY")}");var res = await client.PostAsJsonAsync("https://silkdock.ai/v1/videos/generations", new {model = "seedance-2.0",prompt = "Ocean waves crashing on rocks, sunlight refracting into rainbows",seconds = 10,extra_body = new {resolution = "1080p",aspect_ratio = "16:9",},});Console.WriteLine(await res.Content.ReadAsStringAsync());

require "net/http"require "json"uri = URI("https://silkdock.ai/v1/videos/generations")req = Net::HTTP::Post.new(uri)req["Authorization"] = "Bearer #{ENV['SILKDOCK_API_KEY']}"req["Content-Type"] = "application/json"req.body = {model: "seedance-2.0",prompt: "Ocean waves crashing on rocks, sunlight refracting into rainbows",seconds: 10,extra_body: {resolution: "1080p",aspect_ratio: "16:9",},}.to_jsonres = Net::HTTP.start(uri.host, uri.port, use_ssl: true) { |h| h.request(req) }puts JSON.parse(res.body)["id"]

import java.net.http.*import java.net.URIval req = HttpRequest.newBuilder().uri(URI.create("https://silkdock.ai/v1/videos/generations")).header("Authorization", "Bearer ${System.getenv("SILKDOCK_API_KEY")}").header("Content-Type", "application/json").POST(HttpRequest.BodyPublishers.ofString("""{"model":"seedance-2.0","prompt":"Ocean waves crashing on rocks, sunlight refracting into rainbows","seconds":10,"extra_body":{"resolution":"1080p","aspect_ratio":"16:9"}}""")).build()println(HttpClient.newHttpClient().send(req, HttpResponse.BodyHandlers.ofString()).body())

use reqwest::blocking::Client;use serde_json::json;fn main() -> Result<(), Box<dyn std::error::Error>> {let res = Client::new().post("https://silkdock.ai/v1/videos/generations").header("Authorization", format!("Bearer {}", std::env::var("SILKDOCK_API_KEY")?)).json(&json!({"model": "seedance-2.0","prompt": "Ocean waves crashing on rocks, sunlight refracting into rainbows","seconds": 10,"extra_body": {"resolution": "1080p","aspect_ratio": "16:9"}})).send()?;println!("{}", res.text()?);Ok(())}

POST /v1/videos/generations HTTP/1.1Host: silkdock.aiAuthorization: Bearer <YOUR_API_KEY>Content-Type: application/json{"model": "seedance-2.0","prompt": "Ocean waves crashing on rocks, sunlight refracting into rainbows","seconds": 10,"extra_body": {"resolution": "1080p","aspect_ratio": "16:9"}}

import 'dart:convert';import 'package:http/http.dart' as http;void main() async {final res = await http.post(Uri.parse('https://silkdock.ai/v1/videos/generations'),headers: {'Authorization': 'Bearer ${const String.fromEnvironment("SILKDOCK_API_KEY")}','Content-Type': 'application/json',},body: jsonEncode({'model': 'seedance-2.0','prompt': 'Ocean waves crashing on rocks, sunlight refracting into rainbows','seconds': 10,'extra_body': {'resolution': '1080p','aspect_ratio': '16:9',},}),);final data = jsonDecode(res.body);print(data['id']);}

library(httr2)req <- request("https://silkdock.ai/v1/videos/generations") |>req_headers(Authorization = paste("Bearer", Sys.getenv("SILKDOCK_API_KEY")),"Content-Type" = "application/json") |>req_body_json(list(model = "seedance-2.0",prompt = "Ocean waves crashing on rocks, sunlight refracting into rainbows",seconds = 10,extra_body = list(resolution = "1080p",aspect_ratio = "16:9")))resp <- req_perform(req)cat(resp_body_json(resp)$id)

(* requires cohttp-lwt-unix *)open Cohttp_lwt_unixopen Cohttpopen Lwtlet () =let body = {|{"model":"seedance-2.0","prompt":"Ocean waves crashing on rocks, sunlight refracting into rainbows","seconds":10,"extra*body":{"resolution":"1080p","aspect_ratio":"16:9"}}|} inlet headers = Header.of_list ["Authorization", "Bearer " ^ Sys.getenv "SILKDOCK_API_KEY";"Content-Type", "application/json";] inLwt_main.run (Client.post ~headers ~body:(Cohttp_lwt.Body.of_string body)(Uri.of_string "https://silkdock.ai/v1/videos/generations") >>= fun (*, body) ->Cohttp_lwt.Body.to_string body >>= fun s ->print_string s; return_unit)

curl -X POST https://silkdock.ai/v1/videos/generations \-H "Authorization: Bearer $SILKDOCK_API_KEY" \-H "Content-Type: application/json" \-d '{  "model": "seedance-2.0",  "prompt": "Ocean waves crashing on rocks, sunlight refracting into rainbows",  "seconds": 10,  "extra_body": {    "resolution": "1080p",    "aspect_ratio": "16:9"  }}'

curl -X POST https://silkdock.ai/v1/videos/generations \-H "Authorization: Bearer $SILKDOCK_API_KEY" \-H "Content-Type: application/json" \-d '{  "model": "wan-2.7",  "prompt": "The subject starts running as the camera slowly pulls back",  "input_reference": "https://example.com/reference.jpg",  "seconds": 10,  "extra_body": {    "resolution": "1080p",    "aspect_ratio": "4:3"  }}'

input_reference is passed as a regular reference image and does not automatically become first_frame_url. For first-frame or start/end-frame generation, pass extra_body.first_frame_url and optionally extra_body.last_frame_url explicitly. Check /model_group/info for supports_video_start_end_frame before using this mode. For multimodal reference image, video, or audio lists, check supports_video_multimodal_reference.

curl -X POST https://silkdock.ai/v1/videos/generations \-H "Authorization: Bearer $SILKDOCK_API_KEY" \-H "Content-Type: application/json" \-d '{  "model": "seedance-2.0",  "prompt": "The camera slowly pushes in from the front and ends on a night city silhouette",  "seconds": 6,  "extra_body": {    "resolution": "1080p",    "aspect_ratio": "16:9",    "first_frame_url": "https://example.com/first.png",    "last_frame_url": "https://example.com/last.png"  }}'

curl -X POST https://silkdock.ai/v1/videos/generations \-H "Authorization: Bearer $SILKDOCK_API_KEY" \-H "Content-Type: application/json" \-d '{  "model": "seedance-2.0",  "prompt": "A dancer spins to the beat",  "seconds": 10,  "extra_body": {    "resolution": "720p",    "aspect_ratio": "1:1",    "audio_url": "https://example.com/music.mp3"  }}'

curl -X POST https://silkdock.ai/v1/videos/generations \-H "Authorization: Bearer $SILKDOCK_API_KEY" \-H "Content-Type: application/json" \-d '{  "model": "veo-3.1-generate",  "prompt": "Birdsong echoes through a forest",  "seconds": 8,  "size": "1920x1080"}'

curl -X GET "https://silkdock.ai/model_group/info?model_group=seedance-2.0" \-H "Authorization: Bearer $SILKDOCK_API_KEY"

supports_video_start_end_frame: true means the model supports start/end-frame generation. supports_video_multimodal_reference: true means the model supports multimodal reference asset lists.

curl -X POST https://silkdock.ai/v1/videos/generations \-H "Authorization: Bearer $SILKDOCK_API_KEY" \-F "model=wan-2.7" \-F "prompt=She slowly turns around" \-F "seconds=5" \-F 'extra_body={"resolution":"1080p","aspect_ratio":"16:9"}' \-F "input_reference=@/path/to/image.jpg"

Submit Response

{
  "id": "video_xxxxxx",
  "object": "video",
  "status": "processing",
  "usage": {
    "duration_seconds": 10.0,
    "video_resolution": "1080p",
    "has_audio": false
  }
}

Polling for the Result

Poll GET /v1/videos/{video_id} every few seconds until status is "completed".

# Poll until completedwhile true; doRESULT=$(curl -s https://silkdock.ai/v1/videos/task_abc123 \  -H "Authorization: Bearer $SILKDOCK_API_KEY")STATUS=$(echo $RESULT | jq -r '.status')echo "Status: $STATUS"if [ "$STATUS" = "completed" ] || [ "$STATUS" = "failed" ]; then  echo $RESULT | jq .  breakfisleep 5done

async function pollVideo(videoId) {while (true) {  const res = await fetch(`https://silkdock.ai/v1/videos/${videoId}`, {    headers: { "Authorization": `Bearer ${process.env.SILKDOCK_API_KEY}` },  });  const data = await res.json();  console.log("Status:", data.status, "Progress:", data.progress);  if (data.status === "completed") return data;  if (data.status === "failed") throw new Error(data.error ?? "Generation failed");  await new Promise(r => setTimeout(r, 5000));}}// Full flow: submit then pollconst submit = await fetch("https://silkdock.ai/v1/videos/generations", {method: "POST",headers: {"Authorization": `Bearer ${process.env.SILKDOCK_API_KEY}`,"Content-Type": "application/json",},body: JSON.stringify({model: "seedance-2.0",prompt: "Ocean waves crashing on rocks, sunlight refracting into rainbows",seconds: 10,extra_body: {resolution: "1080p",aspect_ratio: "16:9",},}),});const { id } = await submit.json();const result = await pollVideo(id);console.log("Download:", `https://silkdock.ai/v1/videos/${id}/content`);

import requests, os, timeAPI_KEY = os.getenv("SILKDOCK_API_KEY")BASE = "https://silkdock.ai"# Step 1: Submitres = requests.post(f"{BASE}/v1/videos/generations",headers={"Authorization": f"Bearer {API_KEY}"},json={"model": "seedance-2.0","prompt": "Ocean waves crashing on rocks, sunlight refracting into rainbows","seconds": 10,"extra_body": {"resolution": "1080p","aspect_ratio": "16:9"},},)video_id = res.json()["id"]print(f"Task submitted: {video_id}")# Step 2: Pollwhile True:poll = requests.get(f"{BASE}/v1/videos/{video_id}",headers={"Authorization": f"Bearer {API_KEY}"},)data = poll.json()print(f"Status: {data['status']} Progress: {data.get('progress', '-')}")  if data["status"] == "completed":      print(f"Download: {BASE}/v1/videos/{video_id}/content")      break  if data["status"] == "failed":      raise RuntimeError(data.get("error", "Generation failed"))  time.sleep(5)

package mainimport (  "bytes"  "encoding/json"  "fmt"  "io"  "net/http"  "os"  "time")func main() {apiKey := os.Getenv("SILKDOCK_API_KEY")base := "https://silkdock.ai"  // Step 1: Submit  body, _ := json.Marshal(map[string]any{      "model":   "seedance-2.0",      "prompt":  "Ocean waves crashing on rocks, sunlight refracting into rainbows",      "seconds": 10,      "extra_body": map[string]any{          "resolution":   "1080p",          "aspect_ratio": "16:9",      },  })  req, _ := http.NewRequest("POST", base+"/v1/videos/generations", bytes.NewReader(body))  req.Header.Set("Authorization", "Bearer "+apiKey)  req.Header.Set("Content-Type", "application/json")  resp, _ := http.DefaultClient.Do(req)  raw, _ := io.ReadAll(resp.Body)  resp.Body.Close()  var submit struct{ VideoID string `json:"id"` }  json.Unmarshal(raw, &submit)  fmt.Println("Task submitted:", submit.VideoID)  // Step 2: Poll  for {      req, _ = http.NewRequest("GET", base+"/v1/videos/"+submit.VideoID, nil)      req.Header.Set("Authorization", "Bearer "+apiKey)      resp, _ = http.DefaultClient.Do(req)      raw, _ = io.ReadAll(resp.Body)      resp.Body.Close()      var result struct {          Status   string `json:"status"`          Progress int    `json:"progress"`          Error    object `json:"error"`      }      json.Unmarshal(raw, &result)      fmt.Printf("Status: %s  Progress: %d", result.Status, result.Progress)      if result.Status == "completed" {          fmt.Println("Download:", base+"/v1/videos/"+submit.VideoID+"/content")          return      }      if result.Status == "failed" {          fmt.Println("Failed:", result.Error)          return      }      time.Sleep(5 * time.Second)  }}

Completed Response

{
  "id": "video_xxxxxx",
  "status": "completed",
  "url": "https://cdn.example.com/video.mp4"
}

Failed Response

{
  "id": "video_xxxxxx",
  "status": "failed",
  "error": "Content policy violation: prompt was rejected."
}

Response Fields

Field	Type	Description
`id`	`string`	Unique task identifier. Use this to poll status and download the result.
`object`	`string`	Object type, usually `"video"` on task creation.
`status`	`string`	`"processing"` \| `"completed"` \| `"failed"`
`url`	`string`	URL of the generated video when `status` is `"completed"`.
`usage`	`object`	Billing metadata, including duration, resolution, and audio information.
`error`	`object`	Error message when `status` is `"failed"`.

Errors

Scenario	Result
Unsupported input mode, such as an image-to-video model without `input_reference`	`400 BadRequestError`
Unsupported `resolution` and `aspect_ratio` combination	Falls back to the model default and emits a warning log
Task failure	`status: "failed"` with the reason in `error`
Download before the video is ready	`ValueError: video not ready`

Billing

Video generation is billed by model, duration, resolution, and whether audio is included. The usage field contains billing-related metadata, and billing is handled by the gateway.

Additional Endpoints

Method	Path	Description
`POST`	`/v1/videos/generations`	Submit a video generation task
`GET`	`/v1/videos/{video_id}`	Poll task status
`GET`	`/v1/videos/{video_id}/content`	Download video file (returns `video/mp4`)

Download Video

curl "https://silkdock.ai/v1/videos/{video_id}/content" \
  -H "Authorization: Bearer $SILKDOCK_API_KEY" \
  --output video.mp4

Video Generation

On this page