<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Siray AI</title>
    <description>The latest articles on DEV Community by Siray AI (@sirayb96080876b90914).</description>
    <link>https://dev.to/sirayb96080876b90914</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4251173%2F482ec342-ab2d-4269-a007-e9bf3c73bc9c.png</url>
      <title>DEV Community: Siray AI</title>
      <link>https://dev.to/sirayb96080876b90914</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/sirayb96080876b90914"/>
    <language>en</language>
    <item>
      <title>Seedance 2.5 Flash Spicy API: Pricing &amp; Tested Prompts</title>
      <dc:creator>Siray AI</dc:creator>
      <pubDate>Wed, 18 Nov 2026 07:07:26 +0000</pubDate>
      <link>https://dev.to/sirayb96080876b90914/seedance-25-flash-spicy-api-pricing-tested-prompts-hjf</link>
      <guid>https://dev.to/sirayb96080876b90914/seedance-25-flash-spicy-api-pricing-tested-prompts-hjf</guid>
      <description>&lt;p&gt;The Seedance 2.5 Flash Spicy API gives you ByteDance's Seedance 2.5 Flash video model through Siray AI endpoints intended for mature creative work. It costs &lt;strong&gt;about $0.068 per second at 480p and $0.151 per second at 720p&lt;/strong&gt;, for clips of 4 to 30 seconds with optional audio. It suits fast drafts, high-volume social clips and adult-audience projects.&lt;/p&gt;

&lt;h2&gt;
  
  
  Key takeaways
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Three Spicy endpoints:&lt;/strong&gt; text-to-video, image-to-video (first frame, optional last frame) and reference-to-video with up to 10 images, 5 videos and 5 audio files.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Token-based pricing:&lt;/strong&gt; $0.007 per 1K tokens. A 5-second 720p clip costs about $0.76 and a 30-second 720p clip about $4.54.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Fast in our tests:&lt;/strong&gt; seven clips of 5 to 10 seconds each finished in 72 to 100 seconds from submit to result, including five submitted at the same time.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;480p and 720p only:&lt;/strong&gt; there is no 1080p tier. Use the full Seedance 2.5 when you need higher resolution.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Output moderation still applies:&lt;/strong&gt; one of our SFW prompts (a neon city street) was rejected with a copyright-related policy error, and that task was not billed.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  What is Seedance 2.5 Flash?
&lt;/h2&gt;

&lt;p&gt;Seedance 2.5 Flash is a faster, cheaper tier of ByteDance's Seedance 2.5 video model. ByteDance's cloud unit Volcano Engine announced it on November 12, 2026, and API access opened on November 15, 2026, according to &lt;a href="https://pandaily.com/bytedance-seedance-2-5-flash-faster-lower-cost-video-model-volcano-ark-api" rel="noopener noreferrer"&gt;Pandaily&lt;/a&gt; and other launch coverage. ByteDance says Flash generates video about twice as fast as Seedance 2.5 at roughly 34% lower cost, with output close to the flagship. We could not open ByteDance's own Flash pages to confirm these figures, so treat them as reported vendor claims.&lt;/p&gt;

&lt;p&gt;The flagship &lt;a href="https://seed.bytedance.com/en/seedance2_5" rel="noopener noreferrer"&gt;Seedance 2.5&lt;/a&gt; is described by ByteDance Seed as "a next-generation audio-video joint generation model, built for 30-second storytelling", with up to 30 seconds per generation, more reliable editing and better understanding of reference videos. Flash keeps the 30-second length and native audio but, per launch coverage, drops to 720p and accepts fewer references.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Known limits.&lt;/strong&gt; Flash outputs only 480p and 720p. Siray AI's reference endpoint accepts up to 10 images, 5 videos and 5 audio files, fewer than the flagship's reported 50 references. ByteDance's own speed claims have not been independently benchmarked; our timings below are one data point.&lt;/p&gt;

&lt;h3&gt;
  
  
  What does Spicy mean?
&lt;/h3&gt;

&lt;p&gt;Spicy is a Siray AI catalog label, not a separate ByteDance model. The &lt;a href="https://docs.siray.ai/api-reference/model-api/seedance-2.5-flash-t2v-spicy" rel="noopener noreferrer"&gt;Siray AI API reference&lt;/a&gt; says the variant "supports mature and boundary-pushing creative content for production workflows." On Siray AI, Seedance 2.5 Flash is currently available only as Spicy endpoints.&lt;/p&gt;

&lt;p&gt;Fewer restrictions do not mean no rules. Siray AI's &lt;a href="https://blog.siray.ai/uncensored-ai-generation/" rel="noopener noreferrer"&gt;policy for uncensored generation&lt;/a&gt; says all content "must depict legal adult (18+) individuals only", and illegal content is refused. Don't use a real person's likeness without their consent, and follow your local laws and the rules of the platforms where you publish. Every example in this guide is SFW. For an adult-oriented walkthrough, see &lt;a href="https://blog.siray.ai/seedance-2-5-spicy-slow-dance/" rel="noopener noreferrer"&gt;Seedance 2.5 Spicy Slow Dance: Direct Two Adult Partners&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  Seedance 2.5 Flash Spicy endpoints on Siray AI
&lt;/h2&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Model ID&lt;/th&gt;
&lt;th&gt;Task&lt;/th&gt;
&lt;th&gt;Required inputs&lt;/th&gt;
&lt;th&gt;Optional inputs&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;bytedance/seedance-2.5-flash-t2v-spicy&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Text-to-video&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;prompt&lt;/code&gt;, &lt;code&gt;duration&lt;/code&gt;, &lt;code&gt;size&lt;/code&gt;, &lt;code&gt;aspect_ratio&lt;/code&gt;
&lt;/td&gt;
&lt;td&gt;&lt;code&gt;audio_enable&lt;/code&gt;&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;bytedance/seedance-2.5-flash-i2v-spicy&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Image-to-video&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;prompt&lt;/code&gt;, &lt;code&gt;image&lt;/code&gt;, &lt;code&gt;duration&lt;/code&gt;, &lt;code&gt;size&lt;/code&gt;, &lt;code&gt;aspect_ratio&lt;/code&gt;
&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;end_image&lt;/code&gt;, &lt;code&gt;audio_enable&lt;/code&gt;
&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;bytedance/seedance-2.5-flash-ref2v-spicy&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Reference-to-video, edit, extend&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;prompt&lt;/code&gt;, &lt;code&gt;duration&lt;/code&gt;, &lt;code&gt;size&lt;/code&gt;, &lt;code&gt;aspect_ratio&lt;/code&gt;
&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;images&lt;/code&gt; (≤10), &lt;code&gt;videos&lt;/code&gt; (≤5), &lt;code&gt;audios&lt;/code&gt; (≤5), &lt;code&gt;task_type&lt;/code&gt;, &lt;code&gt;audio_enable&lt;/code&gt;
&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;Shared values on all three endpoints:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;code&gt;duration&lt;/code&gt;: any whole number from 4 to 30. The reference endpoint also accepts &lt;code&gt;-1&lt;/code&gt;, which lets the model choose the length.&lt;/li&gt;
&lt;li&gt;
&lt;code&gt;size&lt;/code&gt;: &lt;code&gt;480p&lt;/code&gt; or &lt;code&gt;720p&lt;/code&gt;.&lt;/li&gt;
&lt;li&gt;
&lt;code&gt;aspect_ratio&lt;/code&gt;: &lt;code&gt;adaptive&lt;/code&gt;, &lt;code&gt;16:9&lt;/code&gt;, &lt;code&gt;4:3&lt;/code&gt;, &lt;code&gt;1:1&lt;/code&gt;, &lt;code&gt;3:4&lt;/code&gt;, &lt;code&gt;9:16&lt;/code&gt; or &lt;code&gt;21:9&lt;/code&gt;.&lt;/li&gt;
&lt;li&gt;
&lt;code&gt;task_type&lt;/code&gt; (reference endpoint only): &lt;code&gt;reference&lt;/code&gt;, &lt;code&gt;auto&lt;/code&gt;, &lt;code&gt;edit&lt;/code&gt; or &lt;code&gt;extend&lt;/code&gt;.&lt;/li&gt;
&lt;li&gt;Images can be URLs or data URLs.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Seedance 2.5 Flash Spicy pricing on Siray AI
&lt;/h2&gt;

&lt;p&gt;All three endpoints are billed by output tokens at &lt;strong&gt;$0.007 per 1K tokens&lt;/strong&gt;. Siray AI's pricing note gives the formula: height × width × (input duration + output duration) × 24 ÷ 1024. In practice that works out to these prices for text-to-video, image-to-video and reference-to-video without video input:&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Duration&lt;/th&gt;
&lt;th&gt;480p&lt;/th&gt;
&lt;th&gt;720p&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;5 seconds&lt;/td&gt;
&lt;td&gt;$0.34&lt;/td&gt;
&lt;td&gt;$0.756&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;10 seconds&lt;/td&gt;
&lt;td&gt;$0.68&lt;/td&gt;
&lt;td&gt;$1.512&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;15 seconds&lt;/td&gt;
&lt;td&gt;$1.02&lt;/td&gt;
&lt;td&gt;$2.268&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;30 seconds&lt;/td&gt;
&lt;td&gt;$2.04&lt;/td&gt;
&lt;td&gt;$4.536&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Per second&lt;/td&gt;
&lt;td&gt;≈ $0.068&lt;/td&gt;
&lt;td&gt;≈ $0.151&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;When you pass reference videos, the rate drops to $0.0043 per 1K tokens, but the input video's length is added to the token count. Siray AI lists a range for these jobs; for example, a 5-second 720p reference job with video input costs $0.84 to $3.25 depending on the input.&lt;/p&gt;

&lt;p&gt;Worked examples:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;A 5-second 720p draft:&lt;/strong&gt; $0.756. Twenty drafts cost $15.12.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;A 15-second 9:16 social clip at 480p:&lt;/strong&gt; $1.02.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;A full 30-second 720p spot:&lt;/strong&gt; $4.536.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Prices were read from the Siray AI pricing API on November 16, 2026, with no discount active. Check the &lt;a href="https://console.siray.ai/model-detail/bytedance/seedance-2.5-flash-t2v-spicy?ref=blog.siray.ai" rel="noopener noreferrer"&gt;Seedance 2.5 Flash t2v Spicy model page&lt;/a&gt; for live prices. In our tests, actual charges matched the formula within about 1%: a 5-second 720p clip used 108,900 tokens ($0.762).&lt;/p&gt;

&lt;h2&gt;
  
  
  Quickstart: call the Seedance 2.5 Flash Spicy API
&lt;/h2&gt;

&lt;p&gt;Generation is asynchronous. Submit a task to get a &lt;code&gt;task_id&lt;/code&gt;, then check it every 3–5 seconds. When the status is &lt;code&gt;SUCCESS&lt;/code&gt;, download the video from &lt;code&gt;data.outputs&lt;/code&gt;.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;1. Submit a task&lt;/strong&gt;&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl -X POST https://api.siray.ai/v1/video/generations \
  -H &lt;span class="s2"&gt;"Authorization: Bearer $SIRAY_API_KEY"&lt;/span&gt; \
  -H &lt;span class="s2"&gt;"Content-Type: application/json"&lt;/span&gt; \
  -d &lt;span class="s1"&gt;'{
    "model": "bytedance/seedance-2.5-flash-t2v-spicy",
    "prompt": "A slow cinematic drone push-in over a foggy pine forest at sunrise...",
    "duration": 5,
    "size": "720p",
    "aspect_ratio": "16:9",
    "audio_enable": true
  }'&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;The response contains the task ID. In our tests it was at the top level of the response (&lt;code&gt;task_id&lt;/code&gt;), not under &lt;code&gt;data&lt;/code&gt; as the API reference shows, so read both places.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;2. Poll until the task finishes&lt;/strong&gt;&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://api.siray.ai/v1/video/generations/&lt;span class="o"&gt;&amp;lt;&lt;/span&gt;task_id&lt;span class="o"&gt;&amp;gt;&lt;/span&gt; \
  -H &lt;span class="s2"&gt;"Authorization: Bearer $SIRAY_API_KEY"&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;&lt;code&gt;status&lt;/code&gt; goes from &lt;code&gt;IN_PROGRESS&lt;/code&gt; to &lt;code&gt;SUCCESS&lt;/code&gt; or &lt;code&gt;FAILURE&lt;/code&gt;. On success, &lt;code&gt;data.outputs&lt;/code&gt; holds the MP4 URL and &lt;code&gt;data.usage.completion_tokens&lt;/code&gt; shows the billed tokens. On failure, check &lt;code&gt;fail_code&lt;/code&gt; and &lt;code&gt;fail_reason&lt;/code&gt;.&lt;/p&gt;

&lt;p&gt;For image-to-video, change &lt;code&gt;model&lt;/code&gt; to &lt;code&gt;bytedance/seedance-2.5-flash-i2v-spicy&lt;/code&gt; and add &lt;code&gt;"image": "&amp;lt;URL or data URL&amp;gt;"&lt;/code&gt;. For reference-to-video, use &lt;code&gt;bytedance/seedance-2.5-flash-ref2v-spicy&lt;/code&gt; with &lt;code&gt;"task_type": "reference"&lt;/code&gt; and &lt;code&gt;"images": [...]&lt;/code&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  Seedance 2.5 Flash prompt examples (tested)
&lt;/h2&gt;

&lt;p&gt;We ran every example below on November 16, 2026 through the Siray AI API. Times are measured from submit to finish on the task record. Costs come from the billed tokens.&lt;/p&gt;

&lt;h3&gt;
  
  
  1. Drone camera move with ambient audio
&lt;/h3&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;A slow cinematic drone push-in over a foggy pine forest at sunrise. The camera starts high above the treetops, glides forward and gradually tilts down to reveal a narrow river winding through the valley, golden light breaking through the mist. Smooth, stable motion, no cuts. Natural ambient sound: soft wind, distant birdsong, the faint rush of water. No people.
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;ul&gt;
&lt;li&gt;Endpoint: &lt;code&gt;t2v-spicy&lt;/code&gt;, 5 s, 720p, 16:9, &lt;code&gt;audio_enable: true&lt;/code&gt;
&lt;/li&gt;
&lt;li&gt;Time: 99 seconds · Cost: $0.762 · Output: 1280×720, 24 fps, with an audio track&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F6kmt6597r33xelk31g5p.jpg" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F6kmt6597r33xelk31g5p.jpg" alt="Aerial view of misty pine forest at sunrise with a river revealed in the valley, video frame made with Seedance 2.5 Flash" width="800" height="450"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;&lt;a href="https://blog.siray.ai/content/media/2026/11/seedance-2-5-flash-drone-forest-720p.mp4" rel="noopener noreferrer"&gt;Watch the clip&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;The push-in and the tilt down to the river came out exactly as written, with no jitter. Spelling out the start point, the motion and the reveal kept the camera move readable in five seconds.&lt;/p&gt;

&lt;h3&gt;
  
  
  2. Character action at 480p
&lt;/h3&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;Medium shot in a sunlit pottery studio. A woman in her early 30s with a linen apron and rolled-up sleeves centers wet clay on a spinning wheel, presses her thumbs into the middle and slowly pulls the walls of a bowl upward. Her hands stay on the clay the whole time; the bowl rises steadily as the wheel turns. Static camera at table height, shallow depth of field, warm afternoon light through a dusty window. Adults only, no children, teenagers, babies or minors.
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;ul&gt;
&lt;li&gt;Endpoint: &lt;code&gt;t2v-spicy&lt;/code&gt;, 8 s, 480p, 16:9, no audio&lt;/li&gt;
&lt;li&gt;Time: 85 seconds · Cost: $0.541 · Output: 854×480, 24 fps&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Funfw6p8ub2nmk49ajtuq.jpg" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Funfw6p8ub2nmk49ajtuq.jpg" alt="A woman shaping clay on a pottery wheel in a sunlit studio, video frame made with Seedance 2.5 Flash" width="800" height="450"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;&lt;a href="https://blog.siray.ai/content/media/2026/11/seedance-2-5-flash-potter-wheel-480p.mp4" rel="noopener noreferrer"&gt;Watch the clip&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Her hands stayed on the clay for all eight seconds, which is where many video models fail. The clay grew into a tall vase rather than a bowl, so name the final shape more firmly if it matters. 480p is clearly softer, but it is enough to check motion before paying for 720p.&lt;/p&gt;

&lt;h3&gt;
  
  
  3. Three-shot sequence with sound
&lt;/h3&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;A three-shot sequence at a small seaside bakery at dawn, natural documentary look.
Shot 1 (wide): a baker in her 50s with grey hair tied back and a flour-dusted blue apron unlocks the wooden door of a whitewashed bakery by the harbor and steps inside.
Shot 2 (close-up): her hands shape a round loaf of dough on a floured wooden table, then score the top with a small blade.
Shot 3 (medium shot): she slides the loaf into a glowing brick oven with a long wooden peel and smiles as warm light falls on her face.
Keep the same woman, apron and bakery in every shot. Sound: seagulls outside, the creak of the door, the crackle of the oven fire. Adults only, no children, teenagers, babies or minors. No text, no logos.
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;ul&gt;
&lt;li&gt;Endpoint: &lt;code&gt;t2v-spicy&lt;/code&gt;, 10 s, 720p, 16:9, &lt;code&gt;audio_enable: true&lt;/code&gt;
&lt;/li&gt;
&lt;li&gt;Time: 100 seconds · Cost: $1.518 · Output: 1280×720, with an audio track&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F9msanf3zrp5v5v89845c.jpg" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F9msanf3zrp5v5v89845c.jpg" alt="A baker in a blue apron unlocking the door of a whitewashed bakery by the harbor at dawn, first shot of a three-shot sequence made with Seedance 2.5 Flash" width="800" height="450"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;&lt;a href="https://blog.siray.ai/content/media/2026/11/seedance-2-5-flash-bakery-multi-shot.mp4" rel="noopener noreferrer"&gt;Watch the clip&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;All three shots appeared in order with clean cuts. The blue apron and grey hair matched across shots. Numbering shots and giving each a framing word (wide, close-up, medium) was enough to get the edit.&lt;/p&gt;

&lt;p&gt;Our first version of this prompt, a man walking through a rainy neon-lit city, failed with &lt;code&gt;OutputVideoSensitiveContentDetected.PolicyViolation&lt;/code&gt; and the message "the output video may be related to copyright restrictions." The prompt was SFW. Our guess is that busy signage triggered the copyright check. No tokens were reported for the failed task.&lt;/p&gt;

&lt;h3&gt;
  
  
  4. Image-to-video: animate a product photo
&lt;/h3&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;Animate this product photo. Steam begins to rise slowly from the mug and curls upward; soft morning light shifts gently across the table. The camera makes a slow, smooth orbit of about 20 degrees to the right around the mug. The mug itself stays still and keeps its exact shape, color and details. No hands, no people, no text.
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;ul&gt;
&lt;li&gt;Endpoint: &lt;code&gt;i2v-spicy&lt;/code&gt;, 5 s, 720p, &lt;code&gt;aspect_ratio: adaptive&lt;/code&gt;, one input image (data URL)&lt;/li&gt;
&lt;li&gt;Time: 72 seconds · Cost: $0.766 · Output: 1178×786&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Ff8f1jwbaw5mmtt2eqvdt.jpg" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Ff8f1jwbaw5mmtt2eqvdt.jpg" alt="A sage-green ceramic mug on a stone block with steam rising, video frame made with Seedance 2.5 Flash image-to-video" width="800" height="534"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;&lt;a href="https://blog.siray.ai/content/media/2026/11/seedance-2-5-flash-i2v-mug-steam.mp4" rel="noopener noreferrer"&gt;Watch the clip&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;The mug kept its shape, glaze and handle exactly, and the steam looked natural. The orbit was much smaller than 20 degrees. With &lt;code&gt;adaptive&lt;/code&gt;, the output followed the input photo's ratio (1178×786) instead of a fixed 1280×720.&lt;/p&gt;

&lt;h3&gt;
  
  
  5. Reference-to-video: place a product in a location
&lt;/h3&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;Use image 1 as the location and image 2 as the product. A slow dolly shot across the bright salt flat from image 1 at golden hour; the ceramic mug from image 2 sits on a small wooden stool in the middle of the salt flat, keeping its exact shape and color. Warm wind moves thin clouds across the sky and the long shadow of the stool stretches toward the camera. Calm, minimal, editorial mood. No people, no text.
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;ul&gt;
&lt;li&gt;Endpoint: &lt;code&gt;ref2v-spicy&lt;/code&gt;, 6 s, 720p, 16:9, &lt;code&gt;task_type: reference&lt;/code&gt;, two input images&lt;/li&gt;
&lt;li&gt;Time: 84 seconds · Cost: $0.914 · Output: 1280×720&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fht2kyllm914m88x9f1qw.jpg" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fht2kyllm914m88x9f1qw.jpg" alt="A green mug on a small wooden stool in the middle of a white salt flat at golden hour, video frame made with Seedance 2.5 Flash reference-to-video" width="800" height="450"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;&lt;a href="https://blog.siray.ai/content/media/2026/11/seedance-2-5-flash-ref2v-salt-flat.mp4" rel="noopener noreferrer"&gt;Watch the clip&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Referring to inputs as "image 1" and "image 2" worked: the model took the location from one image and the mug from the other. The mug's colour and handle matched the reference. Camera movement was again gentler than asked.&lt;/p&gt;

&lt;h2&gt;
  
  
  Prompting tips for Seedance 2.5 Flash
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Ask for more camera motion than you want.&lt;/strong&gt; In both image-driven tests, orbits and dollies came out smaller than written. Pure text-to-video followed camera directions closely.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Number your shots.&lt;/strong&gt; "Shot 1 (wide) … Shot 2 (close-up) …" produced clean cuts in a 10-second clip. Repeat the details that must stay the same in every shot.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Name inputs by position.&lt;/strong&gt; On the reference endpoint, "image 1 as the location, image 2 as the product" was enough to assign roles.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Draft at 480p.&lt;/strong&gt; A 480p clip costs 45% of the 720p price and shows whether motion and timing work.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Keep busy signage out of SFW prompts.&lt;/strong&gt; Neon signs and dense brand-like text may trigger the copyright check, as in our failed test.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;State adult ages and add "no children, teenagers, babies or minors".&lt;/strong&gt; This is required by Siray AI's rules for Spicy endpoints and keeps results predictable.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  When to use Seedance 2.5 Flash (and when not to)
&lt;/h2&gt;

&lt;p&gt;&lt;strong&gt;Use it for:&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Fast drafts and A/B variants before a final render.&lt;/li&gt;
&lt;li&gt;High-volume social clips at 480p or 720p.&lt;/li&gt;
&lt;li&gt;Product clips from a photo or from reference images.&lt;/li&gt;
&lt;li&gt;Mature-themed creative work that needs a Spicy endpoint.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;strong&gt;Look elsewhere when:&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;You need 1080p or higher: use the full Seedance 2.5 or another model that offers 1080p.&lt;/li&gt;
&lt;li&gt;You need more than 10 reference images in one request.&lt;/li&gt;
&lt;li&gt;You need a non-Spicy endpoint for compliance reasons: the plain &lt;code&gt;bytedance/seedance-2.5-t2v&lt;/code&gt;, &lt;code&gt;i2v&lt;/code&gt; and &lt;code&gt;ref2v&lt;/code&gt; endpoints are also on Siray AI.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;For a reference-driven example with the full model, see &lt;a href="https://blog.siray.ai/seedance-2-5-museum-clips-parallax-without-reshaping/" rel="noopener noreferrer"&gt;Seedance 2.5 Museum Clips: Parallax Without Reshaping&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  FAQ
&lt;/h2&gt;

&lt;h3&gt;
  
  
  How much does the Seedance 2.5 Flash Spicy API cost per second?
&lt;/h3&gt;

&lt;p&gt;About $0.068 per second at 480p and $0.151 per second at 720p on Siray AI. Billing is $0.007 per 1K output tokens, so a 5-second 720p clip costs about $0.76 (prices checked November 16, 2026).&lt;/p&gt;

&lt;h3&gt;
  
  
  What is the difference between Seedance 2.5 Flash and Seedance 2.5?
&lt;/h3&gt;

&lt;p&gt;Flash is ByteDance's faster, cheaper tier of Seedance 2.5. ByteDance says it is about twice as fast and about 34% cheaper, but it outputs only 480p or 720p and takes fewer reference files.&lt;/p&gt;

&lt;h3&gt;
  
  
  How long can a Seedance 2.5 Flash video be?
&lt;/h3&gt;

&lt;p&gt;From 4 to 30 seconds, chosen with &lt;code&gt;duration&lt;/code&gt;. The reference-to-video endpoint also accepts &lt;code&gt;-1&lt;/code&gt; to let the model pick the length.&lt;/p&gt;

&lt;h3&gt;
  
  
  Does Seedance 2.5 Flash generate audio?
&lt;/h3&gt;

&lt;p&gt;Yes. Set &lt;code&gt;audio_enable: true&lt;/code&gt;. In our tests, clips with audio cost the same as the token formula predicts for the video alone.&lt;/p&gt;

&lt;h3&gt;
  
  
  Does Seedance 2.5 Flash support image-to-video?
&lt;/h3&gt;

&lt;p&gt;Yes. Use &lt;code&gt;bytedance/seedance-2.5-flash-i2v-spicy&lt;/code&gt; with &lt;code&gt;image&lt;/code&gt; as the first frame and optional &lt;code&gt;end_image&lt;/code&gt; as the last frame. The reference endpoint accepts up to 10 images, 5 videos and 5 audio files.&lt;/p&gt;

&lt;h3&gt;
  
  
  How long does Seedance 2.5 Flash take to generate a video?
&lt;/h3&gt;

&lt;p&gt;In our tests, 5- to 10-second clips took 72 to 100 seconds from submit to finish, and five tasks submitted at the same time did not slow each other down.&lt;/p&gt;

&lt;h2&gt;
  
  
  Try Seedance 2.5 Flash Spicy on Siray AI
&lt;/h2&gt;

&lt;p&gt;Get an API key and run your first clip from the &lt;a href="https://console.siray.ai/model-detail/bytedance/seedance-2.5-flash-t2v-spicy?ref=blog.siray.ai" rel="noopener noreferrer"&gt;Seedance 2.5 Flash t2v Spicy model page&lt;/a&gt;. The same key works for the image-to-video and reference-to-video endpoints and for every other model on &lt;a href="https://console.siray.ai/?ref=blog.siray.ai" rel="noopener noreferrer"&gt;Siray AI&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  Sources
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;ByteDance Seed, &lt;a href="https://seed.bytedance.com/en/seedance2_5" rel="noopener noreferrer"&gt;Seedance 2.5&lt;/a&gt;, accessed November 16, 2026.&lt;/li&gt;
&lt;li&gt;BytePlus, &lt;a href="https://www.byteplus.com/en/blog/dreamina-seedance2-5" rel="noopener noreferrer"&gt;Dreamina Seedance 2.5 announcement&lt;/a&gt;, accessed November 16, 2026.&lt;/li&gt;
&lt;li&gt;Pandaily, &lt;a href="https://pandaily.com/bytedance-seedance-2-5-flash-faster-lower-cost-video-model-volcano-ark-api" rel="noopener noreferrer"&gt;ByteDance Releases Seedance 2.5 Flash&lt;/a&gt;, accessed November 16, 2026.&lt;/li&gt;
&lt;li&gt;Siray AI API reference: &lt;a href="https://docs.siray.ai/api-reference/model-api/seedance-2.5-flash-t2v-spicy" rel="noopener noreferrer"&gt;t2v Spicy&lt;/a&gt;, &lt;a href="https://docs.siray.ai/api-reference/model-api/seedance-2.5-flash-i2v-spicy" rel="noopener noreferrer"&gt;i2v Spicy&lt;/a&gt;, &lt;a href="https://docs.siray.ai/api-reference/model-api/seedance-2.5-flash-ref2v-spicy" rel="noopener noreferrer"&gt;ref2v Spicy&lt;/a&gt;, accessed November 16, 2026.&lt;/li&gt;
&lt;li&gt;Siray AI pricing API, &lt;code&gt;GET /v1/models/pricing/{model}&lt;/code&gt;, November 16, 2026.&lt;/li&gt;
&lt;/ul&gt;




&lt;p&gt;&lt;em&gt;Originally published on the &lt;a href="https://blog.siray.ai/seedance-2-5-flash-spicy-api-pricing-prompts/" rel="noopener noreferrer"&gt;Siray AI blog&lt;/a&gt;. Prices were checked on 2026-11-16 and can change; see the live model page for current pricing.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>ai</category>
      <category>api</category>
      <category>generativeai</category>
      <category>videogeneration</category>
    </item>
    <item>
      <title>Kling 4.0 API: Pricing, Endpoints &amp; Tested Prompts</title>
      <dc:creator>Siray AI</dc:creator>
      <pubDate>Wed, 18 Nov 2026 06:57:22 +0000</pubDate>
      <link>https://dev.to/sirayb96080876b90914/kling-40-api-pricing-endpoints-tested-prompts-29fe</link>
      <guid>https://dev.to/sirayb96080876b90914/kling-40-api-pricing-endpoints-tested-prompts-29fe</guid>
      <description>&lt;p&gt;The Kling 4.0 API gives you Kuaishou's Kling 4.0 video model through eight Siray AI endpoints for text-to-video, image-to-video, reference-to-video and video editing. It costs &lt;strong&gt;about $0.157 per second at 720p and $0.235 per second at 1080p&lt;/strong&gt;, and the faster Kling 4.0 Flash costs $0.094 per second at 720p. It suits clips of 3 to 30 seconds with sound, from product shots to multi-shot scenes.&lt;/p&gt;

&lt;h2&gt;
  
  
  Key takeaways
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Eight endpoints:&lt;/strong&gt; Kling 4.0 and Kling 4.0 Flash each have t2v, i2v, ref2v and video-edit endpoints on Siray AI.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Length and resolution:&lt;/strong&gt; Kling 4.0 makes 3–30 second clips at 720p or 1080p. Flash makes 3–20 second clips at 720p only.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Pricing:&lt;/strong&gt; a 5-second clip costs $0.785 at 720p and $1.175 at 1080p with Kling 4.0, or $0.47 with Flash.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Speed in our tests:&lt;/strong&gt; Kling 4.0 jobs took 106–284 seconds. The same 5-second prompt on Flash took 86 seconds instead of 213.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Audio is always on:&lt;/strong&gt; all six of our test clips came back with a two-channel audio track, and none of the endpoints has an audio switch.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  What is Kling 4.0?
&lt;/h2&gt;

&lt;p&gt;Kling 4.0 is the fourth-generation video model from Kling AI, the video product of Kuaishou Technology. Kling AI &lt;a href="https://kling.ai/blog/kling40-kling-visual-realism-creative-control-introducing-storytelling" rel="noopener noreferrer"&gt;announced it on September 30, 2026&lt;/a&gt;, and Kuaishou's &lt;a href="https://finance.yahoo.com/technology/ai/articles/kling-ai-officially-launches-next-185000472.html" rel="noopener noreferrer"&gt;press release&lt;/a&gt; says it officially launched on October 15, 2026. Kling 4.0 is Kuaishou's model; Siray AI offers it through its unified API.&lt;/p&gt;

&lt;p&gt;According to the official announcement, Kling 4.0:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;"generates videos up to 30 seconds long in a single generation";&lt;/li&gt;
&lt;li&gt;accepts "up to 15 reference assets in a single generation", split into up to 10 images, 5 videos and 7 subjects;&lt;/li&gt;
&lt;li&gt;uses "up to 10 keyframe images to define key visual moments";&lt;/li&gt;
&lt;li&gt;supports "high-quality audio and two-channel stereo" with more accurate lip sync;&lt;/li&gt;
&lt;li&gt;can "make targeted edits to character expressions, body movements, camera positions, camera movements, overall style … while keeping the rest of the scene stable";&lt;/li&gt;
&lt;li&gt;takes prompts of up to 8,000 tokens.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Kling 4.0 Flash is the lighter tier. Kling AI describes it as "built for frequent creation" and says it "combines fast generation with excellent value".&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Known limits.&lt;/strong&gt; Kling AI's announcement labels 10-bit HDR at 4K and video extension up to 2 minutes as "coming soon". Kuaishou's October 16 press release still describes 4K as upcoming. On Siray AI, the top resolution is 1080p, and the i2v endpoint takes a first frame and an optional last frame, not 10 keyframes.&lt;/p&gt;

&lt;h2&gt;
  
  
  Kling 4.0 endpoints on Siray AI
&lt;/h2&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Model ID&lt;/th&gt;
&lt;th&gt;Task&lt;/th&gt;
&lt;th&gt;Required inputs&lt;/th&gt;
&lt;th&gt;Optional inputs&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;kuaishou/kling-4.0-t2v&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Text-to-video&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;prompt&lt;/code&gt;, &lt;code&gt;duration&lt;/code&gt;, &lt;code&gt;size&lt;/code&gt;, &lt;code&gt;aspect_ratio&lt;/code&gt;
&lt;/td&gt;
&lt;td&gt;—&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;kuaishou/kling-4.0-i2v&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Image-to-video&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;prompt&lt;/code&gt;, &lt;code&gt;image&lt;/code&gt;, &lt;code&gt;duration&lt;/code&gt;, &lt;code&gt;size&lt;/code&gt;, &lt;code&gt;aspect_ratio&lt;/code&gt;
&lt;/td&gt;
&lt;td&gt;&lt;code&gt;end_image&lt;/code&gt;&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;kuaishou/kling-4.0-ref2v&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Reference-to-video&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;prompt&lt;/code&gt;, &lt;code&gt;duration&lt;/code&gt;, &lt;code&gt;size&lt;/code&gt;, &lt;code&gt;aspect_ratio&lt;/code&gt;
&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;images&lt;/code&gt; (≤10), &lt;code&gt;videos&lt;/code&gt; (≤5)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;kuaishou/kling-4.0-video-edit&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Edit a video&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;prompt&lt;/code&gt;, &lt;code&gt;video&lt;/code&gt;, &lt;code&gt;duration&lt;/code&gt;, &lt;code&gt;size&lt;/code&gt;, &lt;code&gt;aspect_ratio&lt;/code&gt;
&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;elements&lt;/code&gt;, &lt;code&gt;keep_original_sound&lt;/code&gt;
&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;kuaishou/kling-4.0-flash-t2v&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Text-to-video&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;prompt&lt;/code&gt;, &lt;code&gt;duration&lt;/code&gt;, &lt;code&gt;size&lt;/code&gt;, &lt;code&gt;aspect_ratio&lt;/code&gt;
&lt;/td&gt;
&lt;td&gt;—&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;kuaishou/kling-4.0-flash-i2v&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Image-to-video&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;prompt&lt;/code&gt;, &lt;code&gt;image&lt;/code&gt;, &lt;code&gt;duration&lt;/code&gt;, &lt;code&gt;size&lt;/code&gt;, &lt;code&gt;aspect_ratio&lt;/code&gt;
&lt;/td&gt;
&lt;td&gt;—&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;kuaishou/kling-4.0-flash-ref2v&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Reference-to-video&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;prompt&lt;/code&gt;, &lt;code&gt;duration&lt;/code&gt;, &lt;code&gt;size&lt;/code&gt;, &lt;code&gt;aspect_ratio&lt;/code&gt;
&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;images&lt;/code&gt; (≤10), &lt;code&gt;videos&lt;/code&gt; (≤5)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;code&gt;kuaishou/kling-4.0-flash-video-edit&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Edit a video&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;prompt&lt;/code&gt;, &lt;code&gt;video&lt;/code&gt;, &lt;code&gt;duration&lt;/code&gt;, &lt;code&gt;size&lt;/code&gt;, &lt;code&gt;aspect_ratio&lt;/code&gt;
&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;elements&lt;/code&gt;, &lt;code&gt;keep_original_sound&lt;/code&gt;
&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;Values from the Siray AI API reference:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;code&gt;duration&lt;/code&gt;: any whole number from 3 to 30 on Kling 4.0, and 3 to 20 on Flash.&lt;/li&gt;
&lt;li&gt;
&lt;code&gt;size&lt;/code&gt;: &lt;code&gt;720p&lt;/code&gt; or &lt;code&gt;1080p&lt;/code&gt; on Kling 4.0; &lt;code&gt;720p&lt;/code&gt; only on Flash.&lt;/li&gt;
&lt;li&gt;
&lt;code&gt;aspect_ratio&lt;/code&gt;: &lt;code&gt;auto&lt;/code&gt;, &lt;code&gt;16:9&lt;/code&gt;, &lt;code&gt;9:16&lt;/code&gt;, &lt;code&gt;1:1&lt;/code&gt; or &lt;code&gt;21:9&lt;/code&gt;.&lt;/li&gt;
&lt;li&gt;
&lt;code&gt;image&lt;/code&gt; and &lt;code&gt;images&lt;/code&gt; accept a URL or a data URL. &lt;code&gt;end_image&lt;/code&gt;, &lt;code&gt;video&lt;/code&gt; and &lt;code&gt;videos&lt;/code&gt; are documented as URLs.&lt;/li&gt;
&lt;li&gt;
&lt;code&gt;elements&lt;/code&gt; is a list of characters or objects, each with a &lt;code&gt;frontal_image&lt;/code&gt;, &lt;code&gt;reference_images&lt;/code&gt; and an optional reference &lt;code&gt;video&lt;/code&gt;.&lt;/li&gt;
&lt;li&gt;Flash i2v has no &lt;code&gt;end_image&lt;/code&gt;, so last-frame control needs the full model.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Kling 4.0 pricing on Siray AI
&lt;/h2&gt;

&lt;p&gt;Text-to-video, image-to-video and reference-to-video without a video input are billed per second of output:&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Duration&lt;/th&gt;
&lt;th&gt;Kling 4.0, 720p&lt;/th&gt;
&lt;th&gt;Kling 4.0, 1080p&lt;/th&gt;
&lt;th&gt;Flash, 720p&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;5 seconds&lt;/td&gt;
&lt;td&gt;$0.785&lt;/td&gt;
&lt;td&gt;$1.175&lt;/td&gt;
&lt;td&gt;$0.47&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;10 seconds&lt;/td&gt;
&lt;td&gt;$1.57&lt;/td&gt;
&lt;td&gt;$2.35&lt;/td&gt;
&lt;td&gt;$0.94&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;20 seconds&lt;/td&gt;
&lt;td&gt;$3.14&lt;/td&gt;
&lt;td&gt;$4.70&lt;/td&gt;
&lt;td&gt;$1.88&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;30 seconds&lt;/td&gt;
&lt;td&gt;$4.71&lt;/td&gt;
&lt;td&gt;$7.05&lt;/td&gt;
&lt;td&gt;—&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Per second&lt;/td&gt;
&lt;td&gt;≈ $0.157&lt;/td&gt;
&lt;td&gt;≈ $0.235&lt;/td&gt;
&lt;td&gt;≈ $0.094&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;Jobs that take a video input (video-edit, and ref2v with &lt;code&gt;videos&lt;/code&gt;) bill both the output and the input. Siray AI's pricing note gives the formula: billable duration = output duration + max(input duration, output duration). The per-second rates for these jobs are:&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Endpoint&lt;/th&gt;
&lt;th&gt;Output&lt;/th&gt;
&lt;th&gt;Input video&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;Kling 4.0, 720p&lt;/td&gt;
&lt;td&gt;$0.094/s&lt;/td&gt;
&lt;td&gt;$0.094/s&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Kling 4.0, 1080p&lt;/td&gt;
&lt;td&gt;$0.141/s&lt;/td&gt;
&lt;td&gt;$0.141/s&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Flash, 720p&lt;/td&gt;
&lt;td&gt;$0.063/s&lt;/td&gt;
&lt;td&gt;$0.063/s&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;Worked examples:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;A 5-second 1080p product shot:&lt;/strong&gt; $1.175. Ten takes cost $11.75.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;The same 5 seconds as a Flash draft:&lt;/strong&gt; $0.47, which is 40% less than Kling 4.0 at 720p.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Editing a 5-second clip with Flash:&lt;/strong&gt; 5 seconds out plus 5 seconds in at $0.063 each, $0.63.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;A full 30-second 1080p scene:&lt;/strong&gt; $7.05.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Prices were read from the Siray AI pricing API on November 17, 2026, with no discount shown. Check the &lt;a href="https://console.siray.ai/model-detail/kuaishou/kling-4.0-t2v?ref=blog.siray.ai" rel="noopener noreferrer"&gt;Kling 4.0 t2v model page&lt;/a&gt; for live prices. Failed requests are not billed, according to the same pricing note.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://blog.siray.ai/content/media/2026/11/kling-4-0-concept-duration.mp4" rel="noopener noreferrer"&gt;Watch the clip: a short concept video about how cost grows with clip length, made with Kling 4.0 Flash&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;&lt;em&gt;Concept video generated with Kling 4.0 Flash on Siray AI.&lt;/em&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  Quickstart: call the Kling 4.0 API
&lt;/h2&gt;

&lt;p&gt;Generation is asynchronous. Submit a task to get a &lt;code&gt;task_id&lt;/code&gt;, then check it every 3–5 seconds. When the status is &lt;code&gt;SUCCESS&lt;/code&gt;, download the video from &lt;code&gt;data.outputs&lt;/code&gt;.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;1. Submit a task&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;This is the request we used for our first test:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl -X POST https://api.siray.ai/v1/video/generations \
  -H &lt;span class="s2"&gt;"Authorization: Bearer $SIRAY_API_KEY"&lt;/span&gt; \
  -H &lt;span class="s2"&gt;"Content-Type: application/json"&lt;/span&gt; \
  -d &lt;span class="s1"&gt;'{
    "model": "kuaishou/kling-4.0-t2v",
    "prompt": "A slow, low-angle dolly shot moving forward along a rain-soaked cobblestone alley...",
    "duration": 5,
    "size": "1080p",
    "aspect_ratio": "16:9"
  }'&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;In our tests the task ID came back at the top level of the response (&lt;code&gt;task_id&lt;/code&gt;), not under &lt;code&gt;data&lt;/code&gt; as the API reference shows, so read both places.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;2. Poll until the task finishes&lt;/strong&gt;&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl https://api.siray.ai/v1/video/generations/&lt;span class="o"&gt;&amp;lt;&lt;/span&gt;task_id&lt;span class="o"&gt;&amp;gt;&lt;/span&gt; \
  -H &lt;span class="s2"&gt;"Authorization: Bearer $SIRAY_API_KEY"&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;&lt;code&gt;data.status&lt;/code&gt; moves through &lt;code&gt;NOT_START&lt;/code&gt;, &lt;code&gt;SUBMITTED&lt;/code&gt;, &lt;code&gt;QUEUED&lt;/code&gt; and &lt;code&gt;IN_PROGRESS&lt;/code&gt; to &lt;code&gt;SUCCESS&lt;/code&gt; or &lt;code&gt;FAILURE&lt;/code&gt;. On success, &lt;code&gt;data.outputs&lt;/code&gt; holds the MP4 URL. Two details we noticed: &lt;code&gt;fail_reason&lt;/code&gt; also contains the output URL on successful tasks, so judge success by &lt;code&gt;status&lt;/code&gt; only, and the task record has no usage field for these endpoints.&lt;/p&gt;

&lt;p&gt;To switch tasks, change &lt;code&gt;model&lt;/code&gt; and add the inputs: &lt;code&gt;"image"&lt;/code&gt; (and optionally &lt;code&gt;"end_image"&lt;/code&gt;) for i2v, &lt;code&gt;"images": [...]&lt;/code&gt; for ref2v, or &lt;code&gt;"video": "&amp;lt;URL&amp;gt;"&lt;/code&gt; for video-edit.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://blog.siray.ai/content/media/2026/11/kling-4-0-concept-async-flow.mp4" rel="noopener noreferrer"&gt;Watch the clip: a short concept video about the submit-and-poll flow, made with Kling 4.0 Flash&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;&lt;em&gt;Concept video generated with Kling 4.0 Flash on Siray AI.&lt;/em&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  Kling 4.0 prompt examples (tested)
&lt;/h2&gt;

&lt;p&gt;We ran every example below on November 17, 2026 through the Siray AI API. Times are measured from submit to finish on the task record. Costs are calculated from Siray AI's list prices, because the task record does not report usage for these endpoints. Every clip came back at 24 fps with a two-channel audio track.&lt;/p&gt;

&lt;h3&gt;
  
  
  1. Camera move at 1080p
&lt;/h3&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;A slow, low-angle dolly shot moving forward along a rain-soaked cobblestone alley in a Japanese mountain town at dusk. Paper lanterns glow above wooden shopfronts, steam drifts from a small noodle stall, and puddles reflect the warm light. In the final second the camera rises smoothly above the rooftops to reveal forested mountains under a deep blue sky. Cinematic 35mm film look, soft grain, natural motion. No people. No text, no signs with letters, no logos.
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;ul&gt;
&lt;li&gt;Endpoint: &lt;code&gt;kling-4.0-t2v&lt;/code&gt;, 5 s, 1080p, 16:9&lt;/li&gt;
&lt;li&gt;Time: 213 seconds · Cost: $1.175 · Output: 1920×1080&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F1v8v887p700fj20zmwp6.jpg" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F1v8v887p700fj20zmwp6.jpg" alt="Rain-soaked cobblestone alley in a Japanese mountain town at dusk with lanterns and wet tiled roofs, video frame made with Kling 4.0" width="800" height="450"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;&lt;a href="https://blog.siray.ai/content/media/2026/11/kling-4-0-alley-dolly-1080p.mp4" rel="noopener noreferrer"&gt;Watch the clip&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;The low dolly, the steam, the puddle reflections and the final rise over the rooftops all appear, in that order. Putting the camera move in a sentence of its own ("In the final second the camera rises…") gave the model a clear point to change the motion.&lt;/p&gt;

&lt;h3&gt;
  
  
  2. The same prompt on Kling 4.0 Flash
&lt;/h3&gt;

&lt;ul&gt;
&lt;li&gt;Endpoint: &lt;code&gt;kling-4.0-flash-t2v&lt;/code&gt;, 5 s, 720p, 16:9, same prompt as example 1&lt;/li&gt;
&lt;li&gt;Time: 86 seconds · Cost: $0.47 · Output: 1280×720&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fad6av86zf2bj9oi8y7jg.jpg" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fad6av86zf2bj9oi8y7jg.jpg" alt="The same alley rendered by Kling 4.0 Flash with long rows of round lanterns and a polished stone path, video frame made with Kling 4.0 Flash" width="800" height="450"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;&lt;a href="https://blog.siray.ai/content/media/2026/11/kling-4-0-flash-alley-dolly-720p.mp4" rel="noopener noreferrer"&gt;Watch the clip&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Flash finished in less than half the time and kept the dolly and the rise. The look changed more than the motion did. Flash replaced the scattered lanterns with long, even rows, smoothed the cobblestones and made the scene cleaner and more stylized. It works well for checking timing and camera moves before you pay for a 1080p take.&lt;/p&gt;

&lt;h3&gt;
  
  
  3. Three shots in one 10-second clip
&lt;/h3&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;A three-shot sequence in a small bicycle repair shop on a sunny morning, natural documentary look.
Shot 1 (wide): a mechanic in his 40s with a short grey beard and a green canvas apron rolls up the metal shutter of the shop and wheels a blue bicycle inside.
Shot 2 (close-up): his hands fit a new chain onto the rear sprocket and spin the pedal; the wheel turns smoothly.
Shot 3 (medium shot): he hands the bicycle to a woman in her 30s at the door, she nods, and he wipes his hands on a rag and smiles.
Keep the same man, apron, bicycle and shop in every shot. Adults only, no children, teenagers, babies or minors. No text, no logos.
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;ul&gt;
&lt;li&gt;Endpoint: &lt;code&gt;kling-4.0-t2v&lt;/code&gt;, 10 s, 720p, 16:9&lt;/li&gt;
&lt;li&gt;Time: 236 seconds · Cost: $1.57 · Output: 1280×720&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F547gmb3lh0v97kq0fe1m.jpg" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F547gmb3lh0v97kq0fe1m.jpg" alt="Bearded mechanic in a green apron handing a blue bicycle to a woman at a repair shop door, video frame made with Kling 4.0" width="800" height="450"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;&lt;a href="https://blog.siray.ai/content/media/2026/11/kling-4-0-bike-shop-multi-shot.mp4" rel="noopener noreferrer"&gt;Watch the clip&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Kling 4.0 cut between three distinct shots and kept the same man, apron, blue bicycle and shop throughout. Labelling each shot with its framing (wide, close-up, medium) gave clean cuts. The close-up showed the rear sprocket and chain, though the hands turned the sprocket rather than spinning the pedal.&lt;/p&gt;

&lt;h3&gt;
  
  
  4. Image-to-video with a first and last frame
&lt;/h3&gt;

&lt;p&gt;We gave the i2v endpoint two photos of the same woman on a rooftop. The &lt;code&gt;image&lt;/code&gt; shows her in a moss-green wool coat. The &lt;code&gt;end_image&lt;/code&gt; shows her in just the cream turtleneck and charcoal trousers, with the coat folded on the wall behind her. We made the end frame by editing the first photo with Seedream 5.5, so the face, light and background match exactly.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fra9gisi1mkg984xplg65.jpg" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fra9gisi1mkg984xplg65.jpg" alt="End frame used as end_image: the same woman on the rooftop in a cream turtleneck, her green coat folded on the white wall to her right, image edited with Seedream 5.5" width="800" height="1066"&gt;&lt;/a&gt;&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;The woman on the rooftop slowly unbuttons her moss-green wool coat, slips it off her shoulders, folds it once and lays it on the white parapet wall to her right, leaving her in the cream turtleneck sweater and charcoal trousers. She then turns back toward the camera and settles into the same pose as the end frame. Warm golden-hour light, gentle breeze in her hair, steady handheld camera. Adult woman only, fully clothed. No text, no logos.
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;ul&gt;
&lt;li&gt;Endpoint: &lt;code&gt;kling-4.0-i2v&lt;/code&gt; with &lt;code&gt;end_image&lt;/code&gt;, 5 s, 720p, &lt;code&gt;aspect_ratio: auto&lt;/code&gt;
&lt;/li&gt;
&lt;li&gt;Time: 284 seconds · Cost: $0.785 · Output: 836×1116, matching the shape of the input photos&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fpr6cyms2870jd6qcg499.jpg" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fpr6cyms2870jd6qcg499.jpg" alt="Woman on a Lisbon-style rooftop at golden hour sliding a green wool coat off her shoulders over a cream turtleneck, video frame made with Kling 4.0" width="800" height="1068"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;&lt;a href="https://blog.siray.ai/content/media/2026/11/kling-4-0-i2v-coat-off.mp4" rel="noopener noreferrer"&gt;Watch the clip&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;The clip starts on the first photo and lands on the end frame, with the same face, sweater, rooftop and light. In between, Kling 4.0 played the whole action in order: she unbuttons the coat, slides it off both shoulders, turns away to lay it on the wall and turns back to the camera. The path is approximate, not exact. For about a second the coat stands up stiffly on the wall before it settles into the folded shape from the end frame.&lt;/p&gt;

&lt;p&gt;An earlier version of this test used a navy blazer of the same cut as the end frame. That made the result read as a color swap rather than a coat coming off. &lt;strong&gt;An end frame that differs clearly in shape, not only in color, gives the model a transition it can show.&lt;/strong&gt;&lt;/p&gt;

&lt;h3&gt;
  
  
  5. Reference-to-video: a person and a product
&lt;/h3&gt;

&lt;p&gt;We passed two images to ref2v: a potter in her studio and a stoneware pour-over coffee set on a plain background.&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;Use image 1 as the location and the person, and image 2 as the product. The grey-haired potter from image 1 stops her wheel, wipes her hands on her apron and walks to a wooden table by the rainy window, where the stoneware pour-over coffee set from image 2 is waiting. She pours hot water from the kettle into the dripper in a slow spiral while steam rises in the soft window light. The camera pushes in slowly from a medium shot to a close-up of the coffee dripping into the cup. Keep the shape, glaze and colour of the coffee set exactly as in image 2, and keep the potter's face, hair and apron as in image 1. Adults only. No text, no logos.
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;ul&gt;
&lt;li&gt;Endpoint: &lt;code&gt;kling-4.0-ref2v&lt;/code&gt;, &lt;code&gt;images&lt;/code&gt; with 2 items, 5 s, 720p, 16:9&lt;/li&gt;
&lt;li&gt;Time: 106 seconds · Cost: $0.785 · Output: 1280×720&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fifhwy41bur212t3oottz.jpg" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fifhwy41bur212t3oottz.jpg" alt="Close-up of a gooseneck kettle pouring water into a fluted stoneware dripper with a brass ring, video frame made with Kling 4.0" width="800" height="450"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;&lt;a href="https://blog.siray.ai/content/media/2026/11/kling-4-0-ref2v-potter-coffee.mp4" rel="noopener noreferrer"&gt;Watch the clip&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;The product held up best: the fluted dripper, the brass ring and the spouted pitcher match the reference photo. The studio, the rainy window and the potter's hair and apron also carried over. The opening was compressed: she is already walking with the kettle when the clip begins, so the wheel and the hand wipe are skipped. Numbering the images ("image 1", "image 2") and saying what each is for kept the two references apart.&lt;/p&gt;

&lt;h3&gt;
  
  
  6. Video edit with Flash: rain to snow
&lt;/h3&gt;

&lt;p&gt;We sent the Flash clip from example 2 to the Flash video-edit endpoint.&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;Change the weather from rain to a quiet snowfall. Cover the cobblestones, rooftops and the noodle stall roof with a fresh layer of white snow, replace the puddles with soft snow, and add slow, large snowflakes falling through the lantern light. Keep the same street, buildings, lanterns, steam and camera movement as the original video. No people. No text, no logos.
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;ul&gt;
&lt;li&gt;Endpoint: &lt;code&gt;kling-4.0-flash-video-edit&lt;/code&gt;, 5 s input, 5 s, 720p, 16:9&lt;/li&gt;
&lt;li&gt;Time: 74 seconds · Cost: $0.63 (5 s output + 5 s input) · Output: 1280×720&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F1lw4ft44uspuyn95tydn.jpg" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F1lw4ft44uspuyn95tydn.jpg" alt="The same lantern-lit alley covered in fresh snow with snowy rooftops, video frame made with Kling 4.0 Flash video edit" width="800" height="450"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;&lt;a href="https://blog.siray.ai/content/media/2026/11/kling-4-0-flash-video-edit-snow.mp4" rel="noopener noreferrer"&gt;Watch the clip&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;The street, lanterns, steam, stall and camera move stayed the same, and snow now covers the ground and roofs. One detail went the opposite way: the spots where the puddles were became bare patches of cobblestone in the snow. We could not see falling snowflakes in the frames we checked. This was the fastest job in our tests.&lt;/p&gt;

&lt;h2&gt;
  
  
  Prompting tips for Kling 4.0
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Give each camera move its own sentence.&lt;/strong&gt; "In the final second the camera rises…" produced a clean change of motion at the right moment.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Label shots with their framing.&lt;/strong&gt; "Shot 1 (wide) … Shot 2 (close-up) …" gave three clear cuts in a 10-second clip. Repeat what must stay the same in every shot.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Number reference images and give each a role.&lt;/strong&gt; "Use image 1 as the location and the person, and image 2 as the product" kept the two inputs apart.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Put the key action early.&lt;/strong&gt; In both image-driven tests, the first step of a multi-step action was the one most likely to be skipped or shortened.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Draft on Flash, finish on Kling 4.0.&lt;/strong&gt; The Flash take kept the camera move at 40% of the 720p price, so you can check timing before paying for 1080p.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Plan for sound.&lt;/strong&gt; Every clip comes with audio. Describe the sound you want, or strip the track in editing if you don't need it.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  When to use Kling 4.0 (and when not to)
&lt;/h2&gt;

&lt;p&gt;&lt;strong&gt;Use Kling 4.0 for:&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Final 1080p clips up to 30 seconds with sound.&lt;/li&gt;
&lt;li&gt;Multi-shot scenes where the same person and props must carry across cuts.&lt;/li&gt;
&lt;li&gt;Product videos from reference photos, where the product's shape and finish must hold.&lt;/li&gt;
&lt;li&gt;First-and-last-frame transitions, such as outfit or product reveals.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;strong&gt;Use Kling 4.0 Flash for:&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Drafts, timing checks and A/B variants at 720p.&lt;/li&gt;
&lt;li&gt;High-volume social clips up to 20 seconds.&lt;/li&gt;
&lt;li&gt;Quick video edits, such as changing weather or time of day.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;strong&gt;Look elsewhere when:&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;You need 4K or clips longer than 30 seconds: Kling AI lists 4K and 2-minute extension as coming soon, and Siray AI does not offer them yet.&lt;/li&gt;
&lt;li&gt;You need control over every frame with up to 10 keyframes: the Siray AI i2v endpoint takes a first and a last frame.&lt;/li&gt;
&lt;li&gt;You need a lower price per second at 480p: &lt;a href="https://blog.siray.ai/seedance-2-5-flash-spicy-api-pricing-prompts/" rel="noopener noreferrer"&gt;Seedance 2.5 Flash&lt;/a&gt; costs about $0.068 per second at 480p.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;For more Kling examples, see &lt;a href="https://blog.siray.ai/kling-3-0-omni/" rel="noopener noreferrer"&gt;Kling 3.0 Omni&lt;/a&gt;, &lt;a href="https://blog.siray.ai/kling-3-motion-control/" rel="noopener noreferrer"&gt;Kling 3.0 Motion Control&lt;/a&gt; and the comparison &lt;a href="https://blog.siray.ai/veo-vs-kling-vs-seedance-2-0/" rel="noopener noreferrer"&gt;Veo vs Kling vs Seedance 2.0&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  FAQ
&lt;/h2&gt;

&lt;h3&gt;
  
  
  How much does the Kling 4.0 API cost per second?
&lt;/h3&gt;

&lt;p&gt;About $0.157 per second at 720p and $0.235 per second at 1080p on Siray AI. A 5-second clip costs $0.785 at 720p and $1.175 at 1080p (prices checked November 17, 2026).&lt;/p&gt;

&lt;h3&gt;
  
  
  What is the difference between Kling 4.0 and Kling 4.0 Flash?
&lt;/h3&gt;

&lt;p&gt;Flash is the faster, cheaper tier. On Siray AI it outputs 720p only, makes clips of up to 20 seconds instead of 30, and its i2v endpoint has no last frame. It costs about $0.094 per second, and in our test the same prompt finished in 86 seconds instead of 213.&lt;/p&gt;

&lt;h3&gt;
  
  
  How long can a Kling 4.0 video be?
&lt;/h3&gt;

&lt;p&gt;From 3 to 30 seconds on Kling 4.0 and 3 to 20 seconds on Flash, chosen with &lt;code&gt;duration&lt;/code&gt;.&lt;/p&gt;

&lt;h3&gt;
  
  
  Does Kling 4.0 generate audio?
&lt;/h3&gt;

&lt;p&gt;Yes. Every clip in our tests came with a two-channel audio track. None of the Siray AI endpoints has a switch to turn audio off. On the video-edit endpoints, &lt;code&gt;keep_original_sound&lt;/code&gt; keeps the input video's sound.&lt;/p&gt;

&lt;h3&gt;
  
  
  Can Kling 4.0 edit an existing video?
&lt;/h3&gt;

&lt;p&gt;Yes. Send a video URL to &lt;code&gt;kuaishou/kling-4.0-video-edit&lt;/code&gt; or &lt;code&gt;kuaishou/kling-4.0-flash-video-edit&lt;/code&gt; with a prompt describing the change. You pay for both the output and the input duration. Our 5-second Flash edit cost $0.63.&lt;/p&gt;

&lt;h3&gt;
  
  
  How long does Kling 4.0 take to generate a video?
&lt;/h3&gt;

&lt;p&gt;In our tests, Kling 4.0 took 106 to 284 seconds for 5- to 10-second clips. Flash took 74 to 86 seconds.&lt;/p&gt;

&lt;h2&gt;
  
  
  Try Kling 4.0 on Siray AI
&lt;/h2&gt;

&lt;p&gt;Get an API key and run your first clip from the &lt;a href="https://console.siray.ai/model-detail/kuaishou/kling-4.0-t2v?ref=blog.siray.ai" rel="noopener noreferrer"&gt;Kling 4.0 t2v model page&lt;/a&gt;. The same key works for all eight Kling 4.0 endpoints and every other model on &lt;a href="https://console.siray.ai/?ref=blog.siray.ai" rel="noopener noreferrer"&gt;Siray AI&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  Sources
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;Kling AI, &lt;a href="https://kling.ai/blog/kling40-kling-visual-realism-creative-control-introducing-storytelling" rel="noopener noreferrer"&gt;Kling 4.0: visual realism, creative control and storytelling&lt;/a&gt;, September 30, 2026, accessed November 17, 2026.&lt;/li&gt;
&lt;li&gt;Kuaishou Technology, &lt;a href="https://finance.yahoo.com/technology/ai/articles/kling-ai-officially-launches-next-185000472.html" rel="noopener noreferrer"&gt;Kling AI officially launches next-generation model&lt;/a&gt;, press release, accessed November 17, 2026.&lt;/li&gt;
&lt;li&gt;Siray AI API reference: &lt;a href="https://docs.siray.ai/api-reference/model-api/kling-4.0-t2v" rel="noopener noreferrer"&gt;t2v&lt;/a&gt;, &lt;a href="https://docs.siray.ai/api-reference/model-api/kling-4.0-i2v" rel="noopener noreferrer"&gt;i2v&lt;/a&gt;, &lt;a href="https://docs.siray.ai/api-reference/model-api/kling-4.0-ref2v" rel="noopener noreferrer"&gt;ref2v&lt;/a&gt;, &lt;a href="https://docs.siray.ai/api-reference/model-api/kling-4.0-video-edit" rel="noopener noreferrer"&gt;video-edit&lt;/a&gt;, &lt;a href="https://docs.siray.ai/api-reference/model-api/kling-4.0-flash-t2v" rel="noopener noreferrer"&gt;Flash t2v&lt;/a&gt;, &lt;a href="https://docs.siray.ai/api-reference/model-api/kling-4.0-flash-i2v" rel="noopener noreferrer"&gt;Flash i2v&lt;/a&gt;, &lt;a href="https://docs.siray.ai/api-reference/model-api/kling-4.0-flash-ref2v" rel="noopener noreferrer"&gt;Flash ref2v&lt;/a&gt;, &lt;a href="https://docs.siray.ai/api-reference/model-api/kling-4.0-flash-video-edit" rel="noopener noreferrer"&gt;Flash video-edit&lt;/a&gt;, accessed November 17, 2026.&lt;/li&gt;
&lt;li&gt;Siray AI pricing API, &lt;code&gt;GET /v1/models/pricing/{model}&lt;/code&gt;, November 17, 2026.&lt;/li&gt;
&lt;/ul&gt;




&lt;p&gt;&lt;em&gt;Originally published on the &lt;a href="https://blog.siray.ai/kling-4-0-api-pricing-prompts/" rel="noopener noreferrer"&gt;Siray AI blog&lt;/a&gt;. Prices were checked on 2026-11-17 and can change; see the live model page for current pricing.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>ai</category>
      <category>api</category>
      <category>generativeai</category>
      <category>videogeneration</category>
    </item>
  </channel>
</rss>
