Seedance 2.0 API Pricing (2026): Costs per Second and Code

Seedance 2.0 API pricing comes down to two numbers: how many seconds of video you generate and whether you render at 480p or 720p. On Unicorn API the published price table for both Seedance models keys only on duration and resolution, so the prompt, the aspect ratio and the audio switch do not move the cost. That makes Seedance one of the easier video models to budget for, once you know the per-second rates and how they turn into dollars.

This guide gives you the exact credit cost of every Seedance 2.0 and Seedance 2.0 Fast clip length, what real projects cost on each plan, how to quote a job in code before you run it, which cheaper models are worth a look, and the mistakes that quietly double a video bill.

How Seedance 2.0 API pricing works

Seedance 2.0 is ByteDance’s high-quality video model, released on April 4, 2026. It does text-to-video (you describe the shot in words) or image-to-video (you give it a start frame to animate), at 480p or 720p, for 4 to 15 seconds, with native audio, meaning the model generates sound along with the picture instead of you adding it later. Its sibling, Seedance 2.0 Fast, released the same day, takes exactly the same inputs and runs quicker and cheaper.

Unicorn API bills every model in credits. A credit is a fixed unit of spend:

• Monthly plans give you credits at $0.01 each: Starter is $25, Creator $50, Studio $100 and Pro $200 per month.
• Top-ups cost $0.015 per credit and never expire.

Seedance 2.0 API pricing is charged per output (each video you request) and is a flat rate per second at each resolution:

• Seedance 2.0: 10 credits per second at 480p, 22 credits per second at 720p.
• Seedance 2.0 Fast: 8 credits per second at 480p, 18 credits per second at 720p.

Because the rates are whole numbers, the “rounded up to whole credits” rule never adds anything here: a 7-second 720p Seedance 2.0 clip is exactly 154 credits. The full plan details live on the pricing page.

One practical point before the tables: these rates are what you pay through Unicorn API. Per-second figures you see quoted elsewhere often refer to a different variant, a lower resolution or a different billing unit, so always compare like with like: same model, same resolution, same clip length.

Seedance 2.0 vs Seedance 2.0 Fast: price per clip

Here is the credit cost for the clip lengths people actually use. Every length from 4 to 15 seconds is supported; the ones not shown follow the same per-second rate.

Duration	Seedance 2.0, 480p	Seedance 2.0, 720p	Fast, 480p	Fast, 720p
4 s	40	88	32	72
5 s	50	110	40	90
6 s	60	132	48	108
8 s	80	176	64	144
10 s	100	220	80	180
12 s	120	264	96	216
15 s	150	330	120	270

And the same rates in dollars per second, depending on how you buy credits:

Model and resolution	Credits per second	On a monthly plan	With top-up credits
Seedance 2.0, 480p	10	$0.10	$0.15
Seedance 2.0, 720p	22	$0.22	$0.33
Seedance 2.0 Fast, 480p	8	$0.08	$0.12
Seedance 2.0 Fast, 720p	18	$0.18	$0.27

Three things stand out:

• Resolution is the big lever. 720p costs 2.2x the 480p rate on Seedance 2.0 and 2.25x on Fast. Dropping to 480p for drafts saves more than switching variants.
• Fast is a modest discount. It is 20% cheaper at 480p and about 18% cheaper at 720p. Choose it because it is quicker and good enough for the job, not only to save money.
• Length is linear. There is no discount for longer clips and no penalty for short ones, so a 15-second shot costs exactly three 5-second shots.

What Seedance 2.0 costs for real projects

Per-second rates are easy to read and easy to misjudge, because real work involves retries. Here are four budgets worked through at plan prices ($0.01 per credit).

A single product loop

One 5-second 720p clip with Seedance 2.0 costs 110 credits, or $1.10. If it takes three attempts to get the motion right, the clip costs 330 credits. Do those three attempts at 480p with Fast first (3 × 40 = 120 credits), then one final at 720p (110), and you spend 230 credits instead.

A batch of social ads

Say you need 20 vertical (9:16) ads, 8 seconds each, at 720p with Seedance 2.0. The finals alone cost 20 × 176 = 3,520 credits. If you would otherwise re-roll each final twice at 720p, that is another 7,040 credits. Drafting two variants of each ad with Fast at 480p costs 40 × 64 = 2,560 credits. Total with drafts: 6,080 credits ($60.80), which fits inside the Studio plan’s 10,000 credits with room to spare.

An automated daily clip

A pipeline that posts one 10-second 480p Fast clip per day uses 30 × 80 = 2,400 credits a month. That fits the Starter plan (2,500 credits for $25) almost exactly.

Maximum-length finals

A 15-second 720p Seedance 2.0 clip is 330 credits, the most a single Seedance request can cost. The Pro plan’s 20,000 credits cover 60 of them.

How many clips each plan buys

Plan	Price per month	Credits	5 s 720p Seedance 2.0 clips	5 s 720p Fast clips	5 s 480p Fast clips
Starter	$25	2,500	22	27	62
Creator	$50	5,000	45	55	125
Studio	$100	10,000	90	111	250
Pro	$200	20,000	181	222	500

If your usage is spiky, top-ups make sense because they never expire, but you pay 1.5x the plan rate per credit. For anything you do every month, a plan is the cheaper way to buy Seedance 2.0.

How to check Seedance 2.0 API pricing before you generate

The safest way to handle Seedance 2.0 API pricing in code is to ask for the price first. Unicorn API exposes POST /v1/models/price, which returns the exact cost of a request before anything runs. You send it the same model and input you are about to send to /v1/jobs. This is how you catch a typo that turns a 5-second draft into a 15-second 720p render.

The flow has four steps:

1. Create an API key on the keys page and export it as UNICORN_API_KEY.
2. Quote the request with POST /v1/models/price.
3. Submit it with POST /v1/jobs. Video generation is asynchronous: you get a job id back, not a video.
4. Poll GET /v1/jobs/{id} until status is "succeeded", then read the video URL from outputs[0].url.

Quote a clip with curl

curl -X POST https://api.unicornapi.net/v1/models/price \
  -H "Authorization: Bearer $UNICORN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "bytedance/seedance-2.0",
    "input": {
      "prompt": "A slow drone shot over a rocky coastline at sunset, waves breaking",
      "duration": 5,
      "resolution": "720p",
      "aspect_ratio": "16:9"
    }
  }'

For this request you should see 110 credits.

Quote, generate and download in Python

This script quotes the job, refuses to run it if it costs more than a limit you set, then submits it and waits for the video.

import os
import time
import requests

API = "https://api.unicornapi.net"
HEADERS = {"Authorization": f"Bearer {os.environ['UNICORN_API_KEY']}"}
MAX_WAIT_SECONDS = 600

body = {
    "model": "bytedance/seedance-2.0-fast",
    "input": {
        "prompt": "A slow drone shot over a rocky coastline at sunset, waves breaking",
        "duration": 5,
        "resolution": "480p",
        "aspect_ratio": "16:9",
        "with_audio": True,
    },
}

# 1. Quote the exact price before generating anything
quote = requests.post(f"{API}/v1/models/price", json=body, headers=HEADERS, timeout=30)
quote.raise_for_status()
print("Quoted price:", quote.json())

# 2. Submit the job (each POST is a new, separately billed video)
resp = requests.post(f"{API}/v1/jobs", json=body, headers=HEADERS, timeout=30)
resp.raise_for_status()
job_id = resp.json()["id"]
print("Job submitted:", job_id)

# 3. Poll until the job finishes, with an upper time limit
started = time.time()
while True:
    r = requests.get(f"{API}/v1/jobs/{job_id}", headers=HEADERS, timeout=30)
    r.raise_for_status()
    job = r.json()
    status = job["status"]
    if status == "succeeded":
        print("Video URL:", job["outputs"][0]["url"])
        break
    if status in ("failed", "canceled", "cancelled", "error"):
        raise RuntimeError(f"Job ended with status {status}: {job}")
    if time.time() - started > MAX_WAIT_SECONDS:
        raise TimeoutError(f"Job {job_id} still {status}; check it later instead of resubmitting")
    time.sleep(5)

This draft costs 40 credits. To render the final, change the model to bytedance/seedance-2.0 and the resolution to 720p, and the quote becomes 110. For image-to-video, both Seedance models also accept an image start frame, an end_image, and reference inputs (reference_image_urls, reference_video_urls, reference_audio_urls); the API documentation covers each field. If you work from Cursor or Claude Code, the same calls are available through the MCP integration for agents.

Cheaper alternatives to Seedance 2.0

Seedance is not the cheapest video model per second, and it does not have to be. Its strengths are native audio, a wide set of aspect ratios (16:9, 9:16, 1:1, 4:3, 3:4, 21:9 and adaptive) and rich reference inputs. When you do not need those, other models on the same key can cost a fraction as much. Here is a 10-second clip at 720p across the video models on Unicorn API:

Model	10 s at 720p (credits)	Highest resolution	Input	Notes
Seedance 2.0	220	720p	Text or image	Native audio, reference images, video and audio
Seedance 2.0 Fast	180	720p	Text or image	Same options, quicker and cheaper
Kling 4.0 Flash Keyframe To Video	170	720p	Prompt plus first frame	17 credits per second
Gemini Omni Flash 1.2	90	4k	Prompt, images, first and last frame	4 credits/s at 360p, 10 at 1080p, 22 at 4k
Grok Imagine Video 1.5	45	1080p	Image only	Strong motion, can be busy at peak times, 1 to 15 s
Grok Imagine Video 1.0	45	1080p	Text or image	Clips from 6 to 30 s

How to read this:

• Need 1080p or 4K? On Unicorn API, Seedance stops at 720p. Gemini Omni Flash 1.2 renders 1080p at 10 credits per second, less than half of Seedance 2.0 at 720p, and offers 4k at 22 credits per second, the same rate as Seedance at 720p.
• Animating a still at low cost? Grok Imagine Video 1.5 is about 5x cheaper per second than Seedance 2.0 at 720p. Our guide to 15-second Grok Imagine videos covers it in depth.
• Clips longer than 15 seconds? Grok Imagine Video 1.0 goes to 30 seconds in one request.
• Sound, references and framing control in one call? Seedance is the one to pay for.

Because every model sits behind one key and one bill, a common pattern is to route by need: Grok for cheap motion tests, Seedance for finals with audio, Gemini for high-resolution hero shots. For a wider tour, see our AI video generator API comparison and the image-to-video tool guide, or browse the full model catalog.

Mistakes that make Seedance 2.0 cost more than it should

Most overspending on Seedance has nothing to do with the rate card. It comes from how requests are made.

1. Drafting at 720p. Every exploratory render at 720p costs 2.2x a 480p one. Settle composition and motion at 480p with Fast, then render the winner once.
2. Defaulting to the maximum length. A 15-second clip costs three times a 5-second one. If the edit will use 4 seconds, generate 4 or 5.
3. Resubmitting instead of polling. Each POST /v1/jobs is a new, separately billed video. If your poller times out, look the job up again by id rather than sending the request a second time. The script above raises an error instead of resubmitting for this reason.
4. Sending values the model does not list. Seedance supports durations of 4 to 15 seconds and resolutions of 480p and 720p. Asking for 1080p will not get you 1080p, and the price table has a fallback entry for unlisted combinations. Validate inputs in your code and quote before running.
5. Buying steady volume with top-ups. Top-up credits cost 1.5x plan credits. They are good for occasional spikes; they are an expensive way to fund a monthly pipeline.
6. Comparing headline numbers across providers. A low per-second figure often refers to the Fast variant at 480p. When you compare Seedance 2.0 API pricing between services, match the variant, the resolution and whether audio and reference inputs are included.
7. Using Seedance for every shot. If a shot needs no audio and no references, a cheaper model on the same key may do it for a quarter of the price.

Is there a free Seedance 2.0 API?

People searching for “Seedance 2.0 free” usually want one of two things: to try the model before committing, or to run it at scale without paying. The first is easy and cheap; the second does not really exist for a production workload.

If you see a free Seedance offer anywhere, check what you actually get: which variant, what resolution, how many seconds, how long the queue is, and whether you may use the clips commercially. Discussions on Reddit about Seedance 2.0 pricing tend to land on the same point: the cheapest-looking option is often slow or limited, and quality per dollar matters more than the sticker price.

On Unicorn API, the cheapest real test is a 4-second 480p clip with Seedance 2.0 Fast: 32 credits, or $0.32 on a plan. You can also try both Seedance models in the browser video playground before writing any code, which is the quickest way to see whether the model suits your style.

Choosing your setup

• Just exploring: start in the playground, then run a handful of 4 to 5 second Fast clips at 480p.
• A few finished videos a month: Starter covers 22 five-second 720p Seedance 2.0 clips.
• A content pipeline: Studio or Pro, with drafts on Fast at 480p and finals on Seedance 2.0 at 720p, and a price check on every request.

Seedance 2.0 API pricing rewards discipline more than haggling: pick the right variant, keep drafts small and quote before you render. Do those three things and the bill stays predictable, clip after clip.