HappyHorse Reference-to-Video 1.0: Image to Video Online
Bottom line
HappyHorse Reference-to-Video 1.0 generates a new 4-15 second scene around 1 to 9 reference photos, at 720p or 1080p in five aspect ratios; 5 seconds costs 58 or 99 credits. Upload a starting frame, choose HappyHorse Reference-to-Video 1.0 in the Image to Video tool, and animate it on VdoBloom. The credit price appears before every generation.

HappyHorse Reference-to-Video 1.0 is the original HappyHorse mode for keeping a subject consistent. Upload 1 to 9 photos of a person, product or character, write a prompt of up to 5,000 characters describing the scene, and get a new clip of 4 to 15 seconds at 720p or 1080p in one of five frames: 16:9, 9:16, 1:1, 4:3 or 3:4. A 5-second clip costs 58 credits at 720p or 99 at 1080p; 15 seconds costs 173 or 295.
The newer Reference-to-Video 1.1 takes the same 1-9 references and the same lengths, adds 4:5, 5:4, 21:9 and 9:21, and costs less at every setting. So the reason to choose 1.0 is narrow but real: a character or product series already built on 1.0, where switching versions partway through could change how the subject is rendered. If you are starting a new series, start on 1.1 and stay there.
This mode has no sample clip on the page yet. Below: the spec table, every price next to the 1.1 price, how to keep a 1.0 series consistent, the family table, how it compares with WAN 2.7 and Seedance 2.5 reference workflows, and the caveats.
HappyHorse Reference-to-Video 1.0 specs at a glance
What VdoBloom's backend enforces for happyhorse/reference-to-video. References and limits match 1.1; frames and price do not.
| Spec | HappyHorse Reference-to-Video 1.0 on VdoBloom |
|---|---|
| Inputs | 1 to 9 reference images plus a text prompt (both required) |
| Prompt length | Up to 5,000 characters |
| Clip length | 4 to 15 seconds, any whole second |
| Resolutions | 720p or 1080p |
| Aspect ratios | 16:9, 9:16, 1:1, 4:3, 3:4 |
| Price per clip | 46-173 credits at 720p, 79-295 at 1080p; reference count does not change it |
| Cheapest render | 46 credits: 4 seconds at 720p |
| Content label | MODERATE (standard filtering) |
| Where it runs | Image to Video tab only; not on the effect tabs |
| Failed jobs | Credits refunded automatically |
What a 1.0 reference clip costs, and what staying on 1.0 costs
Reference-to-video 1.0 shares its price list with the other two 1.0 modes, and the number of reference images has no effect on the bill. The last column shows the 1.1 price for the same length at 1080p. Over a series that gap is the price of continuity: six 10-second 1080p episodes cost 1,182 credits on 1.0 and 696 on 1.1. Dollar values use the Lite plan rate ($15 for 300 credits). A one-time 300-credit pack is $19.50 and never expires; plan credits reset each billing cycle.
| Length | 720p | 1080p | Same length on 1.1 at 1080p |
|---|---|---|---|
| 4 s (minimum) | 46 cr ($2.30) | 79 cr ($3.95) | 47 cr |
| 5 s | 58 cr ($2.90) | 99 cr ($4.95) | 58 cr |
| 6 s | 69 cr ($3.45) | 119 cr ($5.95) | 70 cr |
| 7 s | 81 cr ($4.05) | 138 cr ($6.90) | 81 cr |
| 8 s | 92 cr ($4.60) | 158 cr ($7.90) | 93 cr |
| 9 s | 104 cr ($5.20) | 177 cr ($8.85) | 104 cr |
| 10 s | 115 cr ($5.75) | 197 cr ($9.85) | 116 cr |
| 11 s | 127 cr ($6.35) | 217 cr ($10.85) | 127 cr |
| 12 s | 138 cr ($6.90) | 236 cr ($11.80) | 139 cr |
| 13 s | 150 cr ($7.50) | 256 cr ($12.80) | 150 cr |
| 14 s | 161 cr ($8.05) | 275 cr ($13.75) | 162 cr |
| 15 s (maximum) | 173 cr ($8.65) | 295 cr ($14.75) | 173 cr |
Read more: HappyHorse reference-to-video pricing per clip and per series · Credit packs and plans
Keeping a 1.0 series consistent
The habits that keep a subject stable are the same in both versions, but they matter more when you are matching existing 1.0 episodes. Keep an archive of the exact reference images used for the series and reuse them unchanged; swapping in a new photo, even a better one, can shift the subject's look. Save each approved prompt and its settings (length, resolution, aspect ratio) so you can reproduce the style of earlier clips. Change only the parts of the prompt that describe the new scene, action and camera.
If the series is open-ended, consider a one-time comparison: render the next episode on both 1.0 and 1.1 with the same references and prompt at 4 seconds and 720p (46 + 36 credits). If 1.1 holds the subject well enough, moving the rest of the series to it saves credits on every clip and opens the wider frames.
- Archive the reference set and reuse it unchanged.
- Save each approved prompt with its length, resolution and aspect ratio.
- Change only scene, action and camera between episodes.
- Test 1.1 on one episode before deciding to switch versions.
Read more: How to use HappyHorse reference-to-video for consistent characters · Reference-to-video prompt workflows
All six HappyHorse pages compared
Reference-to-video 1.0 is the multi-image entry of the original release: the same 1-9 reference images as 1.1, five frames instead of nine, and the 1.0 price list. All six entries share 4-15 seconds, 720p/1080p and the MODERATE label.
| Page | You start from | Images | Prompt | Aspect ratios | 4 s at 720p | 15 s at 1080p |
|---|---|---|---|---|---|---|
| Text-to-Video 1.1 | A text prompt | None | Required, up to 5,000 characters | 9: 16:9, 9:16, 1:1, 4:3, 3:4, 4:5, 5:4, 21:9, 9:21 | 36 cr | 173 cr |
| Image-to-Video 1.1 | One photo | Exactly 1 | Optional, up to 5,000 characters | No setting; frame the photo | 36 cr | 173 cr |
| Reference-to-Video 1.1 | Reference photos plus a prompt | 1 to 9 | Required, up to 5,000 characters | 9, same set as Text-to-Video 1.1 | 36 cr | 173 cr |
| Text-to-Video 1.0 | A text prompt | None | Required, up to 5,000 characters | 5: 16:9, 9:16, 1:1, 4:3, 3:4 | 46 cr | 295 cr |
| Image-to-Video 1.0 | One photo | Exactly 1 | Optional, up to 5,000 characters | No setting; frame the photo | 46 cr | 295 cr |
| Reference-to-Video 1.0 | Reference photos plus a prompt | 1 to 9 | Required, up to 5,000 characters | 5, same set as Text-to-Video 1.0 | 46 cr | 295 cr |
Read more: HappyHorse Text-to-Video 1.1 · HappyHorse Image-to-Video 1.1 · HappyHorse Reference-to-Video 1.1 · HappyHorse Text-to-Video 1.0 · HappyHorse Image-to-Video 1.0
HappyHorse 1.0 vs WAN 2.7 vs Seedance 2.5 for reference-driven video
WAN 2.7 Reference-to-Video is the cheapest of these per clip but limited to 5 or 10 seconds with no ratio setting; it also accepts reference videos. Seedance 2.5 takes up to nine photos tagged @Image1, @Image2 in the prompt, runs to 30 seconds with audio, and costs more per second. HappyHorse 1.0 offers any length from 4 to 15 seconds in five frames. HappyHorse 1.1 is in the table too, because for a new series it is the more direct upgrade from 1.0 than either of the others. Vidu Q4 Reference also uses the 1-9 image panel in the same tab; its price is shown before you generate.
| Model | Lengths | Aspect ratios | References | 5 s at 720p | Longest clip at 720p |
|---|---|---|---|---|---|
| HappyHorse Reference-to-Video 1.1 | 4-15 s, every second | 9, incl. 21:9 and 9:21 | 1-9 images | 45 cr | 135 cr (15 s) |
| HappyHorse Reference-to-Video 1.0 | 4-15 s, every second | 5 | 1-9 images | 58 cr | 173 cr (15 s) |
| WAN 2.7 Reference-to-Video | 5 or 10 s | No ratio setting | Reference images or videos | 32 cr | 64 cr (10 s) |
| Seedance 2.5 (reference photos) | 4-30 s, every second | 6, incl. 21:9 | Up to 9 photos tagged @Image1, @Image2 | 126 cr | 756 cr (30 s) |
Read more: WAN 2.7 Reference-to-Video model page · Seedance 2.5 model page · Vidu Q4 Reference model page
Running Reference-to-Video 1.0 step by step
Open the Image to Video tab and choose HappyHorse, then Reference-to-Video 1.0. The settings panel switches from a single optional image to a Reference Images (1-9) slot; add your set there. Write the scene prompt, then choose length (4-15 seconds), resolution (720p or 1080p) and one of the five aspect ratios. The credit total appears before you generate, and a failed job is refunded automatically.
This mode is not reachable any other way: it is not on the effect tabs. That is deliberate, since it needs a reference set and a scene prompt that only you can supply.
- Tab: Image to Video, then HappyHorse, then Reference-to-Video 1.0.
- Upload 1-9 references in the Reference Images panel.
- Write the scene prompt (required, up to 5,000 characters).
- Set 4-15 s, 720p or 1080p, and one of five ratios; check the credit total.
Limits and honest caveats
Five frames only, with no 4:5, 5:4, 21:9 or 9:21. It costs more than 1.1 at every setting. You must select it yourself in the Image to Video tab: it is excluded from the one-photo effect tabs. References guide identity but do not guarantee it, and crowded scenes or unseen angles raise the risk of drift. There is no audio setting. The filter does not check likeness rights, so use only images of people you may depict. The 10 free signup credits are below the 46-credit minimum.
What creators use HappyHorse Reference-to-Video 1.0 for
- Continuing a character or mascot series whose earlier episodes were generated on Reference-to-Video 1.0.
- Adding new scenes for a product that already has approved 1.0 clips, keeping the rendering consistent.
- Standard-frame (16:9, 9:16, 1:1, 4:3, 3:4) scenes built around one subject from several reference angles.
- A/B checks of 1.0 against 1.1 on the same reference set before committing a long series to one version.
How to use HappyHorse Reference-to-Video 1.0 on VdoBloom
- 1Open the Image to Video tool and select HappyHorse Reference-to-Video 1.0 from the model picker.
- 2Upload your starting image and describe the motion you want in the prompt.
- 3Choose your duration (4–15s), quality (720p/1080p), aspect ratio, review the credit cost, and generate.
Frequently asked questions
What does HappyHorse Reference-to-Video 1.0 do?
It takes 1 to 9 photos of a subject and a prompt, and generates a new 4-15 second scene featuring that subject at 720p or 1080p, in 16:9, 9:16, 1:1, 4:3 or 3:4. Your photos guide appearance; none of them is used as the opening frame.
How many reference images does it accept?
Between 1 and 9 per generation. More images do not cost more.
How much does it cost?
46 credits for 4 seconds at 720p up to 173 for 15 seconds; 79 to 295 at 1080p. A 6-second 720p clip is 69 credits and a 12-second 1080p clip is 236.
Should I use Reference-to-Video 1.0 or 1.1?
1.1 for anything new: same references and lengths, four more aspect ratios and a lower price (for example 58 vs 99 credits for 5 seconds at 1080p). 1.0 when you are continuing a series that was started on it.
Which aspect ratios are available?
Five: 16:9, 9:16, 1:1, 4:3 and 3:4. For 4:5, 5:4, 21:9 or 9:21, use Reference-to-Video 1.1.
Is a prompt required?
Yes, up to 5,000 characters. The references supply the subject; the prompt supplies the scene, action and camera.
Why can't I find it on the effect tabs?
VdoBloom leaves reference models out of the one-photo effect tabs. Open the Image to Video tab and choose HappyHorse, then Reference-to-Video 1.0.
Can I use photos of real people?
Only people you have the right to depict. The MODERATE filter blocks prohibited content but does not check consent or likeness rights.
What kind of reference photos work best?
Clear, well-lit photos where the subject is not partly hidden, covering the angles your scenes need, such as front, three-quarter and profile. Varied angles usually help more than several copies of the same shot, and reusing the same set for every clip keeps a series consistent.
Is there a free way to try it?
The 10 signup credits do not reach the 46-credit minimum. A one-time 300-credit pack is $19.50 with no subscription.
HappyHorse Reference-to-Video 1.0 guides, prices and comparisons
This page is the HappyHorse Reference-to-Video 1.0 hub. Each guide below goes deeper on one question and links back here.
HappyHorse reference-to-video pricing
Cost per clip and per series.
How to use HappyHorse reference-to-video
Consistent characters from a reference set.
Image-to-video vs reference-to-video
Which one fits your shot.
Reference-to-video vs text-to-video (1.0)
Same price, different job.
Reference-to-video prompt workflows
Prompt patterns that keep a subject stable.
Best AI model for image to video
Where multi-reference models fit.