HappyHorse Text-to-Video 1.0 is VdoBloom’s prompt-only video model that renders 4 to 15 second clips in one-second steps at 720p or 1080p, across five aspect ratios, for 46–173 credits at 720p and 79–295 credits at 1080p. It is the original build of the HappyHorse line, it is rated 4.8 in the VdoBloom catalog, and it sits in the MODERATE content tier. This guide covers exactly what it does well, the real numbers behind every tier, how to write prompts it responds to, and the one situation where you should almost certainly pick a different model instead.

What HappyHorse Text-to-Video is best at

HappyHorse 1.0 takes a text prompt and nothing else. There is no image slot, no reference upload, no first-frame anchor — you describe a scene and it builds the whole thing. That narrows what it is for, and the narrowing is useful.

Its real strength is duration granularity. Most text-to-video models hand you two or three fixed lengths, usually 5 and 10 seconds, and you cut to fit. HappyHorse offers every whole second from 4 to 15. If your edit has a 7-second hole between two beats, you generate 7 seconds rather than generating 10 and trimming three away that you already paid for. Over a long sequence that adds up in both credits and re-render attempts.

The second strength is length itself. A 15-second ceiling gives a camera move room to actually complete — a slow push-in that arrives, a pan that reaches its subject, a two-beat action where something happens and then something follows. Five-second models force you to stitch. HappyHorse lets a single generation carry a whole shot.

The real specs

Straight from the model configuration, with no rounding:

Durations: 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 seconds
Resolutions: 720p and 1080p
Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4
Inputs: text prompt only
Content tier: MODERATE
Catalog rating: 4.8

The five ratios cover widescreen, vertical, square and the two 4:3 orientations. If you need ultrawide 21:9, its 9:21 mirror, or the 4:5 and 5:4 feed ratios, those live on the 1.1 revision — see the section below, because there is more to that choice than framing.

Exact credit costs

Every VdoBloom credit is worth five cents at the Lite tier, where $15 buys 300 credits. That makes the dollar column below a direct conversion, and annual billing lowers it further — the full breakdown is on the pricing page.

Duration	720p credits	720p cost	1080p credits	1080p cost
4s	46	$2.30	79	$3.95
5s	58	$2.90	99	$4.95
6s	69	$3.45	119	$5.95
7s	81	$4.05	138	$6.90
8s	92	$4.60	158	$7.90
9s	104	$5.20	177	$8.85
10s	115	$5.75	197	$9.85
12s	138	$6.90	236	$11.80
15s	173	$8.65	295	$14.75

The 11s, 13s and 14s tiers fall on the same straight line (127, 150 and 161 credits at 720p; 217, 256 and 275 at 1080p). The exact figure always appears in the interface before you confirm a generation, so nothing here is a surprise at run time.

Read this before you pick 1.0

Here is the honest part. HappyHorse 1.1 is cheaper than 1.0 at every single duration, at both resolutions, and it supports more aspect ratios. A 5-second 720p clip is 45 credits on 1.1 versus 58 on 1.0. At 1080p the same clip is 58 credits versus 99 — a 41 percent difference. At 15 seconds and 1080p, 1.1 costs 173 credits where 1.0 costs 295.

Durations and resolutions are identical between them, and both carry the same 4.8 rating and MODERATE tier. So the default recommendation is HappyHorse Text-to-Video 1.1, and the honest case for 1.0 is narrow: you have a prompt library already tuned against its output character and you want that exact behaviour reproduced, or you are matching new shots to footage you generated on 1.0 previously. Those are real reasons. “It is the one I found first” is not.

How to prompt it well

Text-only generation means the prompt carries every decision the model would otherwise read off an image. Four habits that measurably help:

Lead with the subject and the shot, not the mood. Open with what is in frame and how it is framed — medium shot, close-up, wide — then let lighting and atmosphere follow. Adjective-first prompts drift.
Name one camera move, and size the clip to it. A push-in, an orbit, a slow tilt. One move per generation. A 4-second clip cannot complete a full orbit; give that move 10 to 12 seconds or it will look clipped.
Describe motion in verbs, not in nouns. “Steam rises and curls from the cup” produces movement. “A cup with steam” often produces a near-still frame with a drifting camera.
Test at 720p, finish at 1080p. A 5-second 720p test is 58 credits against 99 for the same clip at 1080p. Lock the composition cheaply, then spend once on the final render.

A worked example

Say you need a 9:16 opener for a coffee brand, roughly seven seconds, ending on the product.

Test pass: 7 seconds, 720p, 9:16 — 81 credits ($4.05). Prompt: Medium shot of a ceramic pour-over cone on a dark walnut counter, morning window light from the left, thin stream of water descending, steam curling upward, camera pushes in slowly toward the rim.

What to check: does the push-in arrive by the final second, or is it still travelling? Does the steam move for the whole clip or stall halfway? If the move runs out of room, go to 9 seconds rather than rewriting the prompt.

Final pass: the approved prompt at 1080p, 7 seconds — 138 credits ($6.90). Total spend if one test was enough: 219 credits, about $10.95.

When to pick a different model

HappyHorse 1.0 is not the cheap iteration model on VdoBloom, and pretending otherwise would waste your credits. For rapid volume testing, Wan 3.0 runs a 5-second 720p clip at 32 credits and 1080p at 64 — roughly half of 1.0. If you need the clip to start from a photo or a render you already have, this model cannot do it at all: use HappyHorse Image-to-Video for animating a specific still, or HappyHorse Reference-to-Video when you want one subject to stay recognisable across many generated scenes rather than one picture set in motion. Any workflow that animates a photograph of a real person requires that person’s consent before you upload it.

To run it, open the text-to-video workspace and select HappyHorse Text-to-Video 1.0 from the HappyHorse family, or read the catalog entry on the model page first. New accounts start with 10 free credits and no card, though note that 10 credits will not cover a single HappyHorse clip — use them on a cheaper model to learn the workspace, then come back.

Frequently asked questions

How long can a HappyHorse Text-to-Video clip be?

Up to 15 seconds, selectable at any whole second from 4 upward. There is no preset-only restriction, so 7, 11 or 13 second clips are all valid choices.

What does a 10-second 1080p clip cost?

197 credits, which is $9.85 at the Lite rate of $0.05 per credit. The same clip at 720p is 115 credits, or $5.75.

Can I feed HappyHorse Text-to-Video an image?

No. This variant is text-only. Image-to-Video and Reference-to-Video are separate models in the same family and share the 4–15 second range and 720p/1080p output.

Is version 1.0 or 1.1 the better buy?

1.1, in almost every case. It is cheaper at every duration and resolution, supports nine aspect ratios instead of five, and matches 1.0 on durations, resolutions and rating. Choose 1.0 only when you are deliberately matching earlier 1.0 output.

What content does the MODERATE tier allow?

Standard rules comparable to most mainstream video tools. Everyday creative subjects generate normally; prohibited categories are refused at generation time.

How to Use HappyHorse Text-to-Video: Specs, Prompts and Real Credit Costs