Seedance 2.0: The Complete Guide (2026)

Seedance 2.0 is ByteDance's second-generation AI video model, and it's the workhorse behind a lot of the AI-generated video you've probably already seen: text-to-video, image-to-video, and reference-to-video, all with synchronized audio generated in the same pass. This guide covers what it actually is, what it can do, whether it's free, how pricing works, how to use it, and how it compares to the newer Seedance 2.5. If you're trying to decide whether to build on it or just figure out what all the "Seedance 2.0" search results are about, this is the page for that.

What is Seedance 2.0?

Seedance 2.0 is ByteDance's advanced video generation model, released in February 2026. It brings text-to-video, image-to-video, and reference-to-video together at a single endpoint, and it's designed to produce lifelike, edited-feeling results rather than a single flat take. It's the model behind ByteDance's CapCut and Dreamina apps, and it's also available through third-party APIs for teams building it into their own products.

Key features

Multimodal reference inputs. A single request can combine a text prompt, reference images, a short reference clip, and an audio track, with each input tagged by role so the model knows which one drives style, motion, or sound.

Reference-to-video, multiple ways. You can drive a generation from input images, a short reference clip, or audio references, and combine more than one when a shot needs more than a single cue. The same endpoint also handles plain text-to-video and image-to-video from a static starting image.

Text-to-video and image-to-video, with frame control. Start from a text prompt, or animate a still image as the first frame. You can also supply both a first and last frame when you want precise control over how a clip begins and ends.

Synchronized native audio. Seedance 2.0 generates audio and video jointly, so clips come out with synchronized sound by default instead of arriving silent. That covers music, sound effects, beat-aware cuts, and lip-sync, and you can steer it further with an audio reference.

Director-level camera control. A single call can return multiple shots with natural cuts, so the output feels edited rather than a single continuous take. Camera movement and framing can be directed in plain language.

Motion stability and realism. Seedance 2.0 holds up well through fast action and multi-subject scenes, with more stable motion and more believable physics than earlier models. Reference-driven shots preserve the motion of the source clip closely.

Character consistency. It keeps a character's identity stable across movement, angle changes, and scene transitions, which matters for anything with a recurring face: ads, product spots, serialized content.

Editing and extension. Beyond first-pass generation, Seedance 2.0 supports editing and extending existing clips: modifying an element, replacing something, or continuing a shot from its first and last frames without regenerating the whole thing.

Specs at a glance

Is Seedance 2.0 free?

It depends what you mean by "using" it. If you just want to generate a clip through ByteDance's own consumer apps, CapCut and Dreamina both offer Seedance 2.0 access with some free generations before you hit a paywall or usage cap, which is likely where most of the "is Seedance 2.0 free" searches are coming from.

If you're building something (an app, a pipeline, a product feature) rather than generating one-off clips by hand, you need API access, and that isn't free: it's billed per second of output. What is free is testing it. Apiframe gives new accounts free credits to try the Seedance 2.0 API before you add billing, so you can send real requests and see real output before committing to anything.

Seedance 2.0 pricing

The Seedance 2.0 API is billed per second of generated video, so a longer clip costs proportionally more, and resolution affects the per-second rate. You only pay for successful generations. For exact current rates, check the Seedance 2.0 API page, since per-second pricing is the kind of number worth confirming live rather than trusting a snapshot in an article.

How to use Seedance 2.0 (via API)

Using Seedance 2.0 through an API comes down to four steps: get a key, submit a request, get the result, and use it.

First, create an account and grab an API key. Then send a POST request to the video generation endpoint with your prompt and model: "seedance-2":

curl -X POST https://api.apiframe.ai/v2/videos/generate \
  -H "X-API-Key: afk_your_api_key_here" \
  -H "Content-Type: application/json" \
  -d '{
        "prompt": "a cinematic sunrise over a futuristic cityscape, smooth camera push-in",
        "model": "seedance-2",
        "seedanceParams": {
            "resolution": "720p",
            "generate_audio": false
        }
    }'

That call returns a jobId right away, since generation is asynchronous: a 1080p clip with audio takes real time to render (around two minutes on average). From there, either poll the job endpoint until the status is COMPLETED, or register a webhook and let the result come to you, which is the better pattern for production use.

For image-to-video, add a start_image (and optionally an end_image for precise control over how the clip ends). For reference-driven generation, add reference_image_urls, reference_video_urls, or reference_audio_urls to steer style, motion, or sound. The full Seedance 2.0 API docs cover every parameter in detail.

Seedance 2.0 examples and prompts

A few real prompts that show the range of what Seedance 2.0 can do:

Text-to-video: "A cinematic clip of a chef plating a dessert in a warm restaurant kitchen, slow push-in, soft key light."

Product turntable (image-to-video): "Animate this product photo on a turntable, gentle rotation, studio lighting, subtle reflections, ideal for virtual try-on videos."

Multi-shot transition: "A cinematic clip of a city at dawn, a wide skyline cutting to a close-up of a coffee cup, smooth transition."

Music clip with native audio: "A music clip of a dancer moving in sync with an upbeat track, neon-lit interior, beat-matched cuts, audio generated with the clip."

Reference-to-video: "A tracking shot following a cyclist along a coastal road at golden hour, motion matched to a reference video."

First and last frame control: "The camera continues past a doorway into a sunlit garden," using a doorway shot as the first frame and a garden interior as the last.

You can see these generated (and copy the exact prompts) on the Seedance 2.0 model page.

How Seedance 2.0 compares

Against other current video models, Seedance 2.0's strongest differentiators are physical realism, character consistency, and native audio generated in the same pass rather than added afterward. Compared to Veo 3.1, the two trade wins depending on the shot: Veo tends to edge ahead on raw cinematic polish, Seedance 2.0 on motion stability and multi-subject scenes. Compared to Kling 3.0, Seedance 2.0's multimodal reference system (image, video, and audio combined in one request) is generally more flexible, while Kling has its own strengths in stylized motion. None of these are strictly better across the board, which is a big part of why testing more than one model against your actual use case is worth doing before you commit to a pipeline.

What's next: Seedance 2.5

ByteDance's follow-up model, Seedance 2.5, is a substantial step up: single-pass clips up to 30 seconds instead of 15, up to 50 multimodal references instead of around 12, native 4K instead of 1080p, and region-level local editing. It was announced in June 2026 and is rolling out now. If you're already building on Seedance 2.0 and wondering whether to move up, the Seedance 2.5 complete guide has the full breakdown, including a side-by-side comparison table and current release status.

FAQ

What is Seedance 2.0?

Seedance 2.0 is ByteDance's second-generation multimodal AI video model, released in February 2026. It handles text-to-video, image-to-video, and reference-to-video at a single endpoint, with synchronized native audio.

Is Seedance 2.0 free?

ByteDance's own consumer apps (CapCut, Dreamina) offer some free generations. API access for building your own product isn't free, it's billed per second, though Apiframe gives new accounts free credits to test it first.

How much does the Seedance 2.0 API cost?

It's billed per second of generated video, with resolution affecting the rate. Check the Seedance 2.0 API page for current pricing.

How do I use Seedance 2.0?

Get an API key, send a request to the video generation endpoint with your prompt and model: "seedance-2", then poll for the result or receive it via webhook. See the how-to section above for a working example.

Does it generate audio?

Yes. Audio is generated in the same pass as the video by default, covering music, sound effects, beat-aware cuts, and lip-sync.

How is Seedance 2.0 different from Seedance 2.5?

Seedance 2.5 extends clip length from 15 to 30 seconds in a single pass, expands references from about 12 to up to 50, moves resolution from 1080p to native 4K, and adds region-level local editing. See the Seedance 2.5 guide for the full comparison.