Seedance 2.0 vs Happy Horse 1.1: which 1080p model fits your clip?

Seedance 2.0 vs Happy Horse 1.1 on Makernice: both take up to 9 reference images and render 720p or 1080p up to 15 seconds. Seedance 2.0 adds first/last-frame mode and 21:9, and its audio is always on. Happy Horse 1.1 costs $0.216 per second at 1080p versus $0.4252, and allows 3-second clips.

How do Seedance 2.0 and Happy Horse 1.1 compare?

Source: GET /v1/models and GET /v1/pricing. Catalog facts only: we haven't compared outputs side by side, so this post gives no quality or motion verdict.

How much does a clip cost on each?

Same price in every mode a model supports. Clip price = per-second price × seconds, rounded to the cent; the quote shows the exact amount.

Billed in exact micro-USD; a sum of rounded clip prices can differ by a cent. Live: Pricing or GET https://api.makernice.com/v1/pricing (no key needed).

At 1080p, Happy Horse 1.1 costs about 51% of Seedance 2.0 per second: ten 10-second 1080p clips are $21.60 against $42.528 (totals in exact micro-USD). At 720p the two are within 2%.

Against official list prices, both are above on our price list. Seedance 2.0's official price shown on our model page is $0.151 (720p) and $0.374 (1080p) per second, against our $0.1704 and $0.4252. Happy Horse 1.1 is listed at $0.14 (720p) and $0.18 (1080p) in Alibaba Model Studio's International (Singapore) region, and at the same $0.14 and $0.18 on WaveSpeed and Atlas Cloud, against our $0.168 and $0.216. Alibaba's other regions list lower rates: USD 0.123769 (720p) and USD 0.165026 (1080p) in its Global and Beijing regions.

Quote both models · or price a batch in the cost calculator

Which one supports first/last-frame and 21:9?

Seedance 2.0 only. Its flf mode takes a first frame (image_url) and a last frame (last_image_url), and its text-to-video mode accepts 21:9. Happy Horse 1.1 supports text-to-video, image-to-video and reference-to-video, with 16:9, 9:16, 1:1, 4:3 and 3:4 in text-to-video.

Do both generate audio?

Seedance 2.0's audio is always on and can't be turned off; sending generate_audio returns a 400 before anything is held. For Happy Horse 1.1, audio is not listed in the Makernice catalog, so this post doesn't describe it either way.

When should you pick Happy Horse 1.1?

• 1080p at volume. About half the Seedance 2.0 per-second price at 1080p.
• Short cuts. Clips from 3 seconds, where Seedance 2.0 starts at 4.
• Long prompts within a known limit. Happy Horse 1.1 documents a 2,500-character prompt limit.
• References. Up to 9 reference images, the same as Seedance 2.0.

When should you pick Seedance 2.0?

• Keyframe control. First and last frame (flf) for a shot that must start and end on given images.
• Cinematic 21:9.
• Audio in the same call, always on.
• A draft step in the same family: Seedance 2.0 Fast takes the same request at $0.1027/s (720p). See Seedance 2.0 Fast vs Seedance 2.0.

How do you switch models in one request?

Change model. For Happy Horse 1.1, drop last_image_url (no flf mode) and don't use 21:9. Going the other way, Seedance 2.0 needs at least 4 seconds. Quote first: a field or value a model doesn't support returns 400 invalid_request before anything is held.

  {
-   "model": "seedance-2.0",
+   "model": "happy-horse-1.1",
    "mode": "r2v",
    "resolution": "1080p",
    "prompt": "The two characters from the references share an umbrella on a busy crossing.",
    "reference_images": ["https://example.com/ref-1.jpg", "https://example.com/ref-2.jpg"],
    "duration": 10
  }

import os, requests

API = "https://api.makernice.com/v1"
H = {"Authorization": f"Bearer {os.environ['MAKERNICE_API_KEY']}"}
job = {"mode": "t2v", "resolution": "1080p", "duration": 10, "aspect_ratio": "16:9",
       "prompt": "A street musician plays violin under a bridge at night, slow dolly in."}

for model in ("seedance-2.0", "happy-horse-1.1"):
    q = requests.post(f"{API}/generations/quote", json={**job, "model": model}, headers=H).json()
    print(model, f"${int(q['amount']) / 1e6:.2f}")   # $4.25 · $2.16

Quoting is free and holds nothing; each quote is locked for 12 minutes. Create the generation with the quote_id and an Idempotency-Key, then poll (quickstart).

Where does each model win?

• Happy Horse 1.1: about half the 1080p price, a 3-second minimum, and a documented 2,500-character prompt limit.
• Seedance 2.0: first/last frame, 21:9, and audio that is always on.
• Elsewhere: Happy Horse 1.1 costs less on Alibaba Model Studio (International region $0.14/$0.18 per second; other regions lower), WaveSpeed and Atlas Cloud (USD 0.14/0.18, checked October 22, 2026), and ByteDance's list price for Seedance 2.0 is below ours.
• Output quality: we haven't compared side-by-side clips, so we don't rank them.

Quote both models · or price a batch in the cost calculator

Related

• Models: Seedance 2.0 · Happy Horse 1.1 · Seedance 2.0 Fast · All 16 models
• Comparisons: Seedance 2.0 Fast vs Seedance 2.0 · Best models for video remakes
• Guides: Reference-to-video guide
• Cost: AI video cost calculator · No charge for failed generations · Pricing

About Makernice

Makernice is an AI video and image generation API: one API key and one prepaid balance for 16 models from ByteDance, Kuaishou, MiniMax and Alibaba. Every run starts with a quote that locks the price for 12 minutes, through Makernice's own async REST API (quote, create, poll, cancel).