Video models
Seedance, Kling, Hailuo, Veo, and every other frontier video model behind one integration, with the same request shape every time.
Showing all 13 models
by ByteDance
Next-gen ByteDance video — 30s single-pass clips, up to 50 multimodal references (30 images / 10 videos / 10 audio), native synced audio.
by MiniMax
Native 2K video with stereo audio, multimodal reference input, and instruction-based editing.
by Black Forest Labs
Black Forest Labs' multimodal model — up to 20s of video with synchronized audio from text, images, or a clip.
by Alibaba
Happy Horse 1.1 — text-to-video, image-to-video, and reference-to-video (up to 9 images) with synced audio.
by Kuaishou
Latest Kling — per-second pricing, optional audio.
by Alibaba
Wan 2.7 unified text-to-video and image-to-video, with optional FLF and audio sync.
by xAI
xAI video from Grok Imagine.
by Midjourney
Animate a Midjourney still into a short clip.
by Luma
Luma Ray cinematography with concept controls.
by Runway
Latest Runway Gen-4.5 video.
by Google
Google's flagship cinematic video model with audio.
by OpenAI
OpenAI's Sora video generator.
by Topaz
Pro-grade video upscale and frame interpolation.
Missing a model you need? New models land within days of release. Tell us what to add next.
Quickstart
Same request shape as every other modality: prompt in, hosted video out through an async job and a webhook.
Read the docsimport { Apiframe } from "@apiframe-ai/sdk"
const client = new Apiframe ({ apiKey : process.env.APIFRAME_API_KEY })
// submit a job; poll or use a webhook for the result
const { jobId } = await client. videos .generate( {
model: "seedance-2.5" ,
prompt: "aerial shot of a coastal village at golden hour, waves rolling in" ,
seedanceParams: { duration: 10 } ,
})
const job = await client.jobs.waitFor(jobId)
console.log(job.result)
The short version of what teams ask before shipping video features. Everything else lives in the docs.
Most models generate clips of 5-15 seconds, Flux Video goes up to 20 seconds with synchronized audio, and Seedance 2.5 goes up to 30 seconds. Longer sequences are usually stitched from multiple generations.
Yes. Seedance, Veo, Kling, and most of the catalog accept a reference image alongside the prompt to control the first frame or the overall look.
Yes. Models like Seedance, Veo, Kling, and Grok Imagine generate natively synchronized audio. Silent models can be paired with a music model in the same API.
9,800+
developers & companies
Official partner of
Instant access to 70+ media models through a single API. Start free and scale when you're ready.
