freevideoai-h3-tools 0.1.0 → 0.1.2
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +27 -39
- package/guides/first-last-frame.md +4 -4
- package/guides/getting-started.md +9 -9
- package/guides/h3-product-images.md +19 -44
- package/guides/image-to-video.md +2 -2
- package/guides/offline-tools.md +16 -12
- package/guides/text-to-video.md +3 -3
- package/package.json +2 -2
package/README.md
CHANGED
|
@@ -1,10 +1,10 @@
|
|
|
1
1
|
# FreeVideo AI — H3 Tools & Tutorials
|
|
2
2
|
|
|
3
|
-
**
|
|
3
|
+
**FreeVideo AI** is a browser-based AI video generation platform. Its official website is [freevideoai.app](https://freevideoai.app/). The online studio creates short videos from a prompt, a photo, or first and last frames.
|
|
4
4
|
|
|
5
|
-
[
|
|
5
|
+
[Video Studio](https://freevideoai.app/ai-video-generator) · [Pricing](https://freevideoai.app/pricing) · [Product Transparency](https://freevideoai.app/product-transparency) · [H3 Image Preview](https://freevideoai.app/ai-image-generator)
|
|
6
6
|
|
|
7
|
-
This
|
|
7
|
+
This repository contains FreeVideo AI's offline prompt tools and tutorials. FreeVideo AI hosts the online studio, and MiniMax H3 is a model used in its video workflows. Cloud tasks may be processed by third-party generation services. See [Product Transparency](https://freevideoai.app/product-transparency) for workflow and data-handling details. The repository's MIT license covers the offline package and its accompanying materials.
|
|
8
8
|
|
|
9
9
|
## Offline prompt and validation tools
|
|
10
10
|
|
|
@@ -34,22 +34,21 @@ node bin/cli.js validate image examples/image.json
|
|
|
34
34
|
node bin/cli.js validate video examples/video.json
|
|
35
35
|
```
|
|
36
36
|
|
|
37
|
-
Copy
|
|
37
|
+
Copy a video prompt into the website when you are ready to generate. The two example files above are offline validation fixtures. A valid image configuration is suitable for planning while H3 Image remains in Preview and public generation is closed.
|
|
38
38
|
|
|
39
39
|
Read the [CLI and JavaScript tutorial](guides/offline-tools.md) for installation, programmatic usage, defaults and error handling.
|
|
40
40
|
|
|
41
|
-
##
|
|
41
|
+
## Video generation and credits
|
|
42
42
|
|
|
43
|
-
| Starting point
|
|
44
|
-
|
|
|
45
|
-
| A written scene
|
|
46
|
-
| One photo or illustration | Image to Video
|
|
47
|
-
| Two endpoint images
|
|
48
|
-
| A description of an image | H3 Image | 0 credits during the introductory period |
|
|
43
|
+
| Starting point | Mode | H3 Standard cost |
|
|
44
|
+
| ------------------------- | ------------------ | ------------------------ |
|
|
45
|
+
| A written scene | Text to Video | 6 credits per generation |
|
|
46
|
+
| One photo or illustration | Image to Video | 8 credits per generation |
|
|
47
|
+
| Two endpoint images | First & Last Frame | 8 credits per generation |
|
|
49
48
|
|
|
50
|
-
|
|
49
|
+
Eligible new Google accounts receive **60 one-time welcome credits**, valid for **30 days**. At H3 Standard rates, 60 credits cover up to ten Text to Video tasks with no credits left, or seven Image to Video or First & Last Frame tasks with four credits left.
|
|
51
50
|
|
|
52
|
-
|
|
51
|
+
Cloud uses dynamic pricing by model, resolution and duration. The studio shows the quote before submission. Cloud requires eligible purchased credits. Welcome credits apply to H3 Standard only. The [Pricing page](https://freevideoai.app/pricing) lists credit packs and monthly and annual plans. Annual plans save 30% compared with 12 monthly payments.
|
|
53
52
|
|
|
54
53
|
## Make your first video
|
|
55
54
|
|
|
@@ -67,46 +66,35 @@ Start with one subject, one action and one camera move. Review the result, then
|
|
|
67
66
|
|
|
68
67
|
## Tutorials
|
|
69
68
|
|
|
70
|
-
| Guide
|
|
71
|
-
|
|
|
72
|
-
| [Getting started](guides/getting-started.md)
|
|
73
|
-
| [Text to Video](guides/text-to-video.md)
|
|
74
|
-
| [Image to Video](guides/image-to-video.md)
|
|
75
|
-
| [First & Last Frame](guides/first-last-frame.md)
|
|
76
|
-
| [
|
|
69
|
+
| Guide | What you will learn |
|
|
70
|
+
| ----------------------------------------------------- | -------------------------------------------------------------------------------------------- |
|
|
71
|
+
| [Getting started](guides/getting-started.md) | Choose a mode, understand credits, preview and download |
|
|
72
|
+
| [Text to Video](guides/text-to-video.md) | Write single-shot prompts and troubleshoot motion |
|
|
73
|
+
| [Image to Video](guides/image-to-video.md) | Prepare a product photo or portrait and guide subtle movement |
|
|
74
|
+
| [First & Last Frame](guides/first-last-frame.md) | Select compatible images and build a transition |
|
|
75
|
+
| [Product-image concepts](guides/h3-product-images.md) | Draft product prompts while H3 Image is in Preview and use an existing photo in Video Studio |
|
|
77
76
|
|
|
78
|
-
##
|
|
77
|
+
## Plan a product-image concept
|
|
79
78
|
|
|
80
|
-
|
|
79
|
+
[H3 Image](https://freevideoai.app/ai-image-generator) is in **Preview**. Public image generation is currently closed. Use the offline toolkit to prepare a prompt for a tall product poster or a wider product card.
|
|
81
80
|
|
|
82
81
|
```text
|
|
83
82
|
An unbranded skincare serum bottle with clear glass, an ivory inner container, a white pump and brushed gold accents. The bottle floats at a slight diagonal above a glossy black pedestal. Suspended water droplets, a charcoal studio background, warm rim lighting and realistic reflections. One bottle, no lettering or logos.
|
|
84
83
|
```
|
|
85
84
|
|
|
86
|
-
|
|
85
|
+
When reviewing a concept from any image workflow, inspect shapes, labels and reflections. Use an approved product photograph when the output must depict the exact packaging.
|
|
87
86
|
|
|
88
87
|
## Current controls and limits
|
|
89
88
|
|
|
90
|
-
|
|
91
|
-
| --- | --- | --- |
|
|
92
|
-
| Output | Silent MP4 | JPEG, PNG or WebP |
|
|
93
|
-
| Duration | 5, 10 or 15 seconds | Not applicable |
|
|
94
|
-
| Ratios | 16:9, 9:16 | 1:1, 16:9, 9:16, 4:3, 3:4 |
|
|
95
|
-
| Resolution control | Workflow native resolution | 1K or 2K area levels; check actual result dimensions |
|
|
96
|
-
| Reference inputs | One image or two frames in their respective modes | Text only |
|
|
97
|
-
| Upload types | PNG, JPEG, WebP; up to 10 MB per image | No reference-image upload |
|
|
98
|
-
|
|
99
|
-
Image ratios do not add video output ratios. To animate a generated image, download it and upload it manually into Image to Video; that video task costs 8 credits. A PNG output does not imply transparent-background support.
|
|
100
|
-
|
|
101
|
-
The public video workflow does not generate audio or lip sync. Exact typography, product identity, complex motion and endpoint matching are not guaranteed. See [product transparency](https://freevideoai.app/product-transparency) for practical limitations.
|
|
89
|
+
H3 Standard offers 5-, 10-, or 15-second MP4 clips in landscape 16:9 or portrait 9:16. For Image to Video and First & Last Frame, upload PNG, JPEG or WebP files up to 10 MB each. Cloud options vary by model. Exact typography, product identity, complex motion and endpoint matching can vary. See [Product Transparency](https://freevideoai.app/product-transparency) for practical limitations.
|
|
102
90
|
|
|
103
91
|
## Common questions
|
|
104
92
|
|
|
105
|
-
**
|
|
93
|
+
**Can I start for free?** The one-time welcome grant lets eligible new Google accounts try H3 Standard. H3 Image is in Preview and public image generation is closed.
|
|
106
94
|
|
|
107
95
|
**Do I need to install a model or supply an API key?** No for the website: sign in and use the browser interface. This repository does not provide a public generation API or local model installer.
|
|
108
96
|
|
|
109
|
-
**
|
|
97
|
+
**Does Standard include native audio?** H3 Standard MP4s retain model-generated audio, including when using welcome credits. Cloud audio options depend on the model.
|
|
110
98
|
|
|
111
99
|
**Why does a task remain queued or show a result check?** Queue time varies, and the service may still be reconciling a result. Wait for a final status before submitting a duplicate task.
|
|
112
100
|
|
|
@@ -116,7 +104,7 @@ The public video workflow does not generate audio or lip sync. Exact typography,
|
|
|
116
104
|
|
|
117
105
|
- [Live video examples](https://freevideoai.app/examples)
|
|
118
106
|
- [Website video prompt guide](https://freevideoai.app/ai-video-prompt-guide)
|
|
119
|
-
- [
|
|
107
|
+
- [Official creative guides](https://freevideoai.app/guides)
|
|
120
108
|
|
|
121
|
-
Last reviewed
|
|
109
|
+
Last reviewed November 21, 2026. Refer to the website for current controls, availability and pricing.
|
|
122
110
|
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
# First & Last Frame Video: Build a Controlled Transition in FreeVideo AI
|
|
2
2
|
|
|
3
|
-
First & Last Frame uses two uploaded images to guide a video's beginning and ending. Use it
|
|
3
|
+
First & Last Frame uses two uploaded images to guide a video's beginning and ending. Use it for a clear before-and-after composition, such as a product photographed under two lighting setups. In H3 Standard it costs 8 credits per task and exports an MP4. Cloud quotes vary by model, resolution and duration and require eligible purchased credits.
|
|
4
4
|
|
|
5
5
|
[Open First & Last Frame](https://freevideoai.app/ai-video-generator?mode=first-last-frame)
|
|
6
6
|
|
|
@@ -20,7 +20,7 @@ Very different objects or camera viewpoints can lead to warping or unexpected in
|
|
|
20
20
|
|
|
21
21
|
Use PNG, JPEG or WebP, with each file no larger than 10 MB. Keep the subject's scale and placement similar. For video output, choose 16:9 or 9:16; prepare both images close to that orientation.
|
|
22
22
|
|
|
23
|
-
|
|
23
|
+
Use two images you have permission to upload. H3 Image remains in Preview and public image generation is closed.
|
|
24
24
|
|
|
25
25
|
## Generate the transition
|
|
26
26
|
|
|
@@ -63,11 +63,11 @@ No. The images guide a generative workflow. It is not a guarantee that every pix
|
|
|
63
63
|
This mode uses two still images. It is not a public video-to-video editing control.
|
|
64
64
|
|
|
65
65
|
**Will the result include sound?**
|
|
66
|
-
|
|
66
|
+
H3 Standard MP4s retain model-generated audio, including when using welcome credits. Cloud audio options depend on the selected model.
|
|
67
67
|
|
|
68
68
|
See the website's [First & Last Frame guide](https://freevideoai.app/first-last-frame-guide) and [FAQ](https://freevideoai.app/faq) for current product details.
|
|
69
69
|
|
|
70
70
|
---
|
|
71
71
|
|
|
72
|
-
Last reviewed
|
|
72
|
+
Last reviewed November 21, 2026. Example prompts illustrate a starting point. [Back to the tutorial index](../README.md).
|
|
73
73
|
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
# Getting Started with FreeVideo AI: Your First Clip, Credits & Downloads
|
|
2
2
|
|
|
3
|
-
FreeVideo AI turns text or still images into short
|
|
3
|
+
[FreeVideo AI](https://freevideoai.app/) turns text or still images into short MP4 videos. This guide covers a first clip in H3 Standard.
|
|
4
4
|
|
|
5
5
|
## 1. Pick a mode based on what you already have
|
|
6
6
|
|
|
@@ -14,11 +14,11 @@ Open the [video generator](https://freevideoai.app/ai-video-generator). Sign in
|
|
|
14
14
|
|
|
15
15
|
## 2. Check credits before starting
|
|
16
16
|
|
|
17
|
-
Text to Video costs 6 credits per generation. Image to Video and First & Last Frame cost 8 credits each. Eligible new Google accounts receive 60 welcome credits,
|
|
17
|
+
In H3 Standard, Text to Video costs 6 credits per generation. Image to Video and First & Last Frame cost 8 credits each. Eligible new Google accounts receive a one-time grant of 60 welcome credits, valid for 30 days.
|
|
18
18
|
|
|
19
|
-
|
|
19
|
+
With the full 60 credits, you can submit ten Text to Video tasks and use all 60 credits, or seven Image to Video or First & Last Frame tasks and have four credits left.
|
|
20
20
|
|
|
21
|
-
|
|
21
|
+
Cloud quotes depend on the model, resolution and duration. Check the exact quote before submitting. Cloud accepts eligible purchased credits only. See [Pricing](https://freevideoai.app/pricing) for one-time packs and subscriptions. Annual plans save 30% compared with 12 monthly payments.
|
|
22
22
|
|
|
23
23
|
## 3. Make a simple first clip
|
|
24
24
|
|
|
@@ -39,25 +39,25 @@ The prompt gives the model one subject, one visible action and one camera moveme
|
|
|
39
39
|
|
|
40
40
|
Watch the full clip once for the subject, once for movement and once for unwanted changes. If the camera moves too much, revise only the camera instruction. If the background distracts, simplify the setting. Changing one variable makes comparisons easier.
|
|
41
41
|
|
|
42
|
-
|
|
42
|
+
H3 Standard MP4s retain model-generated audio, including when using welcome credits. Cloud audio support depends on the model.
|
|
43
43
|
|
|
44
44
|
## Settings to know
|
|
45
45
|
|
|
46
|
-
|
|
46
|
+
H3 Standard offers 5, 10 or 15 seconds and landscape 16:9 or portrait 9:16. Its workflow determines the output resolution. Start with a short test to establish the look.
|
|
47
47
|
|
|
48
|
-
H3 Image
|
|
48
|
+
[H3 Image](https://freevideoai.app/ai-image-generator) is in Preview and public generation is closed. Choose the video aspect ratio independently in Video Studio.
|
|
49
49
|
|
|
50
50
|
## If something looks stuck
|
|
51
51
|
|
|
52
52
|
- **Queued or generating:** allow the active task to finish; avoid repeatedly submitting the same prompt.
|
|
53
53
|
- **Checking the result:** the service may still be reconciling the task. Review its final status before retrying.
|
|
54
54
|
- **History is missing:** check that you are using the same account, then refresh the history.
|
|
55
|
-
- **No sound:**
|
|
55
|
+
- **No sound:** check the player volume and the audio in the downloaded file. Older clips may have been exported without an audio track.
|
|
56
56
|
- **Unexpected output:** simplify the request; generation is not an exact simulation or guaranteed reproduction.
|
|
57
57
|
|
|
58
58
|
Continue with the [video prompt guide](https://freevideoai.app/ai-video-prompt-guide) or read the [product FAQ](https://freevideoai.app/faq).
|
|
59
59
|
|
|
60
60
|
---
|
|
61
61
|
|
|
62
|
-
Last reviewed
|
|
62
|
+
Last reviewed November 21, 2026. Example prompts illustrate a starting point. [Back to the tutorial index](../README.md).
|
|
63
63
|
|
|
@@ -1,69 +1,44 @@
|
|
|
1
|
-
#
|
|
1
|
+
# Plan Product-Style Images While H3 Image Is in Preview
|
|
2
2
|
|
|
3
|
-
H3 Image in FreeVideo AI
|
|
3
|
+
H3 Image is in Preview on [FreeVideo AI](https://freevideoai.app/). Public image generation is currently closed. You can prepare a product concept offline or animate a photograph you already have in Video Studio.
|
|
4
4
|
|
|
5
|
-
[Open H3 Image](https://freevideoai.app/ai-image-generator)
|
|
5
|
+
[Open H3 Image Preview](https://freevideoai.app/ai-image-generator)
|
|
6
6
|
|
|
7
|
-
##
|
|
7
|
+
## Plan the composition
|
|
8
8
|
|
|
9
|
-
|
|
9
|
+
Use a text prompt to describe a still-image concept, including the product shape, materials, lighting, background and framing. Check an offline plan with `node bin/cli.js validate image examples/image.json` from the repository root. The example is a configuration fixture for the validator.
|
|
10
10
|
|
|
11
|
-
|
|
11
|
+
A rendered concept can differ from a real product's cap, proportions or label. For a listing that must show the exact item, use approved product photography and an editing workflow that preserves it.
|
|
12
12
|
|
|
13
|
-
|
|
13
|
+
For a wider product card, sketch a **4:3** layout. For a taller poster, sketch a **3:4** layout. Compare the placement of the subject and the empty space for copy in each composition. These ratios are also accepted by the offline image validator.
|
|
14
14
|
|
|
15
|
-
|
|
16
|
-
| --- | --- | --- |
|
|
17
|
-
| Aspect ratio | 1:1, 16:9, 9:16, 4:3, 3:4 | Match the placement where the image will appear |
|
|
18
|
-
| Area level | 1K or 2K | These are area levels, not a fixed long-edge pixel count |
|
|
19
|
-
| File format | JPEG, PNG or WebP | Choose the format your destination accepts |
|
|
20
|
-
| Prompt | Up to 2,000 characters | Keep the subject and composition clear |
|
|
21
|
-
|
|
22
|
-
For a product card, try **4:3**. For a taller product poster, try **3:4**. Use 16:9 for a wide hero image or 9:16 for a tall composition. Check the actual pixel dimensions of the result. Selecting PNG does not add a transparent-background feature.
|
|
23
|
-
|
|
24
|
-
## Generate a luxury skincare concept
|
|
25
|
-
|
|
26
|
-
1. Sign in to H3 Image.
|
|
27
|
-
2. Choose **3:4** for a tall bottle composition.
|
|
28
|
-
3. Select an area level and output format.
|
|
29
|
-
4. Enter the prompt below.
|
|
30
|
-
5. Generate one image and wait for its final status.
|
|
31
|
-
6. Inspect the preview, then download the image.
|
|
15
|
+
## Draft a skincare concept prompt
|
|
32
16
|
|
|
33
17
|
```text
|
|
34
18
|
A premium skincare serum bottle with clear glass walls, an ivory inner container, a white pump and brushed gold metal accents. The unbranded bottle floats at a slight diagonal above a glossy black cylindrical pedestal. Suspended water droplets surround it. A charcoal gray studio background, warm rim lighting, soft side light, realistic glass reflections and a subtle pedestal reflection. One main bottle, carefully separated from the background, elegant product photography. No lettering or logos.
|
|
35
19
|
```
|
|
36
20
|
|
|
37
|
-
|
|
21
|
+
The prompt describes an imagined product. When reviewing a render, inspect the water droplets, glass reflections and bottle shape for impossible details or duplicates.
|
|
38
22
|
|
|
39
23
|
For **4:3**, keep the same subject but add: “A wider composition with the bottle slightly right of center and clear negative space on the left.” This leaves room to add accurate copy later.
|
|
40
24
|
|
|
41
|
-
##
|
|
25
|
+
## Review branding in an editor
|
|
42
26
|
|
|
43
|
-
Requesting exact brand text can produce misspellings or distorted lettering. For reliable advertising copy,
|
|
27
|
+
Requesting exact brand text can produce misspellings or distorted lettering. For reliable advertising copy, add approved text or a logo in an image editor.
|
|
44
28
|
|
|
45
29
|
Before publishing, review the pump, cap, bottle outline, shadows and reflections. Check that the visual does not imply an ingredient, certification or product feature that the actual product does not have.
|
|
46
30
|
|
|
47
|
-
##
|
|
48
|
-
|
|
49
|
-
1. Download the image.
|
|
50
|
-
2. Open [Image to Video](https://freevideoai.app/ai-video-generator?mode=image-to-video).
|
|
51
|
-
3. Upload the downloaded PNG, JPEG or WebP, within the 10 MB video-input limit.
|
|
52
|
-
4. Choose the video orientation: **16:9 or 9:16**.
|
|
53
|
-
5. Ask for a small camera move and restrained subject motion.
|
|
54
|
-
6. Generate and inspect the silent MP4.
|
|
55
|
-
|
|
56
|
-
This is a manual download-and-upload workflow. There is no public one-click transfer described here. A 3:4 or 4:3 source will not create a 3:4 or 4:3 video output; prepare a matching video composition first if cropping would be a problem. Image to Video costs 8 credits even while the image tool's introductory generation price is 0.
|
|
57
|
-
|
|
58
|
-
## If a task is waiting
|
|
59
|
-
|
|
60
|
-
Only one active image task per account is supported. Keep the page or history available and wait for the final status instead of submitting duplicates. Queue capacity and service availability can vary. “Checking the result” means the task may still be reconciled.
|
|
31
|
+
## Animate an existing product photo
|
|
61
32
|
|
|
62
|
-
|
|
33
|
+
1. Open [Image to Video](https://freevideoai.app/ai-video-generator?mode=image-to-video) with a PNG, JPEG or WebP photo you have permission to use, up to 10 MB.
|
|
34
|
+
2. For H3 Standard, choose 16:9 or 9:16 and check the framing.
|
|
35
|
+
3. Ask for a small camera move and restrained subject motion.
|
|
36
|
+
4. Review the quote before submitting. H3 Standard Image to Video costs 8 credits per task.
|
|
37
|
+
5. Preview the finished MP4 and inspect the packaging and label before publishing.
|
|
63
38
|
|
|
64
|
-
|
|
39
|
+
Prepare a crop that fits the intended video frame. Read the [product video guide](https://freevideoai.app/product-video-guide) for more examples, and check the [FAQ](https://freevideoai.app/faq) for current availability.
|
|
65
40
|
|
|
66
41
|
---
|
|
67
42
|
|
|
68
|
-
Last reviewed
|
|
43
|
+
Last reviewed November 21, 2026. Example prompts illustrate a starting point. [Back to the tutorial index](../README.md).
|
|
69
44
|
|
package/guides/image-to-video.md
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
# Image to Video with FreeVideo AI: Animate a Product Photo or Portrait
|
|
2
2
|
|
|
3
|
-
Image to Video starts from a still image and uses a written prompt to guide movement.
|
|
3
|
+
Image to Video starts from a still image and uses a written prompt to guide movement. Try it for a product reveal, a portrait animation or a small camera move across an illustration. In H3 Standard it costs 8 credits per task and exports an MP4. H3 Standard MP4s retain model-generated audio, including when using welcome credits. Cloud quotes vary by model, resolution and duration and require eligible purchased credits.
|
|
4
4
|
|
|
5
5
|
[Open Image to Video](https://freevideoai.app/ai-video-generator?mode=image-to-video)
|
|
6
6
|
|
|
@@ -63,5 +63,5 @@ If you need to guide both endpoints, use [First & Last Frame](https://freevideoa
|
|
|
63
63
|
|
|
64
64
|
---
|
|
65
65
|
|
|
66
|
-
Last reviewed
|
|
66
|
+
Last reviewed November 21, 2026. Example prompts illustrate a starting point. [Back to the tutorial index](../README.md).
|
|
67
67
|
|
package/guides/offline-tools.md
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
# FreeVideo AI H3 Tools: CLI & JavaScript Tutorial
|
|
2
2
|
|
|
3
|
-
Prepare
|
|
3
|
+
Prepare prompts and check parameter syntax offline with this package. Use the official [FreeVideo AI website](https://freevideoai.app/) for browser-based video generation.
|
|
4
4
|
|
|
5
5
|
## Install from npm or run from source
|
|
6
6
|
|
|
@@ -62,7 +62,7 @@ node bin/cli.js validate image image.json
|
|
|
62
62
|
|
|
63
63
|
Valid output contains `ok: true` and normalized `value`. Invalid output contains `ok: false` and an `errors` array with field names and explanations.
|
|
64
64
|
|
|
65
|
-
Image defaults are 1:1, 1K and PNG.
|
|
65
|
+
Image defaults in this package are 1:1, 1K and PNG. The offline validator accepts 1:1, 16:9, 9:16, 4:3 and 3:4, the 1K and 2K area levels, and the `jpeg`, `png` and `webp` format values. These fields describe an offline plan. See [H3 Image Preview](https://freevideoai.app/ai-image-generator) for the current website status.
|
|
66
66
|
|
|
67
67
|
## Validate a video plan
|
|
68
68
|
|
|
@@ -76,11 +76,11 @@ Image defaults are 1:1, 1K and PNG. Supported ratios are 1:1, 16:9, 9:16, 4:3 an
|
|
|
76
76
|
}
|
|
77
77
|
```
|
|
78
78
|
|
|
79
|
-
|
|
79
|
+
For a Standard video plan, the validator accepts numeric durations 5, 10 and 15, and ratios 16:9 and 9:16. Text to Video needs zero references, Image to Video one, and First & Last Frame two. `referenceCount` records the planned number of source images. Choose Cloud settings and review the actual quote on the website.
|
|
80
80
|
|
|
81
81
|
Video defaults are text-to-video, 5 seconds, 16:9 and zero references. Unknown keys are rejected to catch misspellings or unsupported capabilities.
|
|
82
82
|
|
|
83
|
-
|
|
83
|
+
The offline validators use a 2,000-character limit after trimming. Images count Unicode code points and video counts UTF-16 code units, so some emoji count as two video units. Validation checks the settings accepted by this package.
|
|
84
84
|
|
|
85
85
|
## Use JavaScript
|
|
86
86
|
|
|
@@ -90,10 +90,12 @@ import {
|
|
|
90
90
|
listTemplates,
|
|
91
91
|
validateImageOptions,
|
|
92
92
|
validateVideoOptions,
|
|
93
|
-
} from
|
|
93
|
+
} from "freevideoai-h3-tools";
|
|
94
94
|
|
|
95
|
-
const prompt = createPrompt(
|
|
96
|
-
|
|
95
|
+
const prompt = createPrompt("product", {
|
|
96
|
+
subject: "An unbranded perfume bottle",
|
|
97
|
+
});
|
|
98
|
+
const result = validateImageOptions({ prompt, aspectRatio: "4:3" });
|
|
97
99
|
if (result.ok) {
|
|
98
100
|
console.log(result.value);
|
|
99
101
|
} else {
|
|
@@ -101,7 +103,9 @@ if (result.ok) {
|
|
|
101
103
|
}
|
|
102
104
|
|
|
103
105
|
console.log(listTemplates());
|
|
104
|
-
console.log(
|
|
106
|
+
console.log(
|
|
107
|
+
validateVideoOptions({ prompt: "A cup with rising steam.", duration: 5 }),
|
|
108
|
+
);
|
|
105
109
|
```
|
|
106
110
|
|
|
107
111
|
When running directly inside a cloned repository, import from `./index.js` instead. TypeScript declarations are included. `createPrompt` throws `TypeError` for unknown templates or invalid overrides; validators return a result object for ordinary JSON inputs.
|
|
@@ -110,10 +114,10 @@ When running directly inside a cloned repository, import from `./index.js` inste
|
|
|
110
114
|
|
|
111
115
|
Use `-` as the file argument to read JSON from standard input. Validation writes JSON to stdout. Exit codes: 0 valid, 1 invalid configuration, 2 malformed JSON, file errors or command errors. Diagnostics for command/file errors go to stderr. Prompts are read only from arguments or the file/stdin you explicitly provide.
|
|
112
116
|
|
|
113
|
-
##
|
|
117
|
+
## Use a video prompt on the website
|
|
114
118
|
|
|
115
|
-
After
|
|
119
|
+
After validating a video plan, open [Video Studio](https://freevideoai.app/ai-video-generator), sign in, paste the prompt and select the settings in the browser. The image example is for offline planning while H3 Image remains in Preview and public generation is closed.
|
|
116
120
|
|
|
117
|
-
|
|
121
|
+
Eligible new Google accounts receive 60 one-time welcome credits, valid for 30 days. They apply to H3 Standard. Cloud quotes are based on model, resolution and duration, and require eligible purchased credits. Review [Pricing](https://freevideoai.app/pricing) and [FAQ](https://freevideoai.app/faq) for current terms.
|
|
118
122
|
|
|
119
|
-
Last reviewed
|
|
123
|
+
Last reviewed November 21, 2026. [Back to README](../README.md).
|
package/guides/text-to-video.md
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
# Text to Video with FreeVideo AI: Write Better Single-Shot Prompts
|
|
2
2
|
|
|
3
|
-
Use Text to Video when you have a scene idea
|
|
3
|
+
Use Text to Video when you have a scene idea without a starting image. FreeVideo AI's H3 Standard workflow uses MiniMax H3 to create a short MP4 from your description at 6 credits per task. Cloud quotes vary by model, resolution and duration and require eligible purchased credits.
|
|
4
4
|
|
|
5
5
|
[Open Text to Video](https://freevideoai.app/ai-video-generator?mode=text-to-video)
|
|
6
6
|
|
|
@@ -61,11 +61,11 @@ Choose a longer duration when a simple movement needs more time, such as a gentl
|
|
|
61
61
|
|
|
62
62
|
## What this workflow does not do
|
|
63
63
|
|
|
64
|
-
|
|
64
|
+
H3 Standard MP4s retain model-generated audio, including when using welcome credits. Exact brand lettering, stable small details and complex physical interactions can be unreliable. Inspect the downloaded result before using it in a product listing or campaign.
|
|
65
65
|
|
|
66
66
|
For current controls and limitations, see [Text to Video](https://freevideoai.app/text-to-video), [FAQ](https://freevideoai.app/faq) and [Product transparency](https://freevideoai.app/product-transparency).
|
|
67
67
|
|
|
68
68
|
---
|
|
69
69
|
|
|
70
|
-
Last reviewed
|
|
70
|
+
Last reviewed November 21, 2026. Example prompts illustrate a starting point. [Back to the tutorial index](../README.md).
|
|
71
71
|
|
package/package.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "freevideoai-h3-tools",
|
|
3
|
-
"version": "0.1.
|
|
4
|
-
"description": "Offline
|
|
3
|
+
"version": "0.1.2",
|
|
4
|
+
"description": "Offline prompt templates and parameter validation with guides for the FreeVideo AI video studio at freevideoai.app.",
|
|
5
5
|
"type": "module",
|
|
6
6
|
"exports": {
|
|
7
7
|
".": {
|