Vikit.ai SDK let you develop easily video generators leveraging generative AI and other AI models. You may see this as a langchain to orchestrate AI models and video editing tools. To see the video in full resolution, click on the GIF.

You need a personal access token, which you can easily obtain from here.
Vikit.ai SDK is easy to test standalone using Google Colab. It is easy to develop through Dev Containers. Dev container file and instructions will be published soon.

• Python 3.8+
• requirements.txt contains requirements for python environment
• FFMPEG and FFPROBE

Generating a video from simple text prompt, with background music and voice over narration:

from vikit.music_building_context import MusicBuildingContext
from vikit.prompt.prompt_factory import PromptFactory
from vikit.video.prompt_based_video import PromptBasedVideo
from vikit.video.video_build_settings import VideoBuildSettings

prompt = "Paris, the City of Light, is a global center of art, fashion, and culture, renowned for its iconic landmarks and romantic atmosphere. The Eiffel Tower, Louvre Museum, and Notre-Dame Cathedral are just a few of the city's must-see attractions."

video_build_settings = VideoBuildSettings(
    music_building_context=MusicBuildingContext(
        apply_background_music=True,
        generate_background_music=True,
    ),
    include_read_aloud_prompt=True,
    test_mode=False,
)

gw = video_build_settings.get_ml_models_gateway()
prompt = await PromptFactory(ml_gateway=gw).create_prompt_from_text(prompt)
video = PromptBasedVideo(prompt=prompt)
await video.build(build_settings=video_build_settings)

You can orchestrate several videos, using CompositeVideo(). Here is an example showing how to generate two videos from texts:

from vikit.video.composite_video import CompositeVideo
from vikit.video.raw_text_based_video import RawTextBasedVideo
from vikit.video.seine_transition import SeineTransition
from vikit.video.video_build_settings import VideoBuildSettings

prompt1 = "Paris, the City of Light, is a global center of art, fashion, and culture, renowned for its iconic landmarks and romantic atmosphere."
prompt2 = "The Eiffel Tower, Louvre Museum, and Notre-Dame Cathedral are just a few of the city's must-see attractions."

video_composite = CompositeVideo()

video1 = RawTextBasedVideo(prompt1)
video2 = RawTextBasedVideo(prompt2)
# Generate the transition to pass from the first to the second video
transit = SeineTransition(
    source_video=video1,
    target_video=video2,
)

video_composite.append_video(video1).append_video(transit).append_video(video2)

await video_composite.build(
    build_settings=VideoBuildSettings(
        test_mode=False,
        output_video_file_name="Composite.mp4",
    )
)

More elaborated examples can be found in script_example.py or in our Google Colab. For additional information, please refer to the Documentation.

If you've encountered a bug, have a feature request, or have any suggestions to improve the project, we encourage you to open an issue on our GitHub repository. To do so, go to the "Issues" tab in our GitHub repository and click "New Issue." Please provide a clear title and detailed description, including steps to reproduce, environment setup, and any relevant screenshots or code snippets. If you have questions, reach out to us hello@vikit.ai.

Wanna help? You're very welcome! A typical Pull Request workflow would be like this:

1. New PR: You submit your PR.
2. PR validation: If the PR passes all the quality checks then Vikit.ai team assign a reviewer. We may ask for additional changes to make to PR pass quality checks.
3. PR review: If everything looks good, the reviewer(s) will approve the PR. The reviewers might ask some modifications before approving your PR.
4. CI tests & Merge: Once the PR is approved we launch CI tests. We may ask further modifications in this step, in order to get all the tests passed before merging your PR. Once all the tests pass, vikit team merge the code internally as well as externally on GitHub.

• Fork vikit repository into your own GitHub account.
• Create a new branch and make your changes to the code.
• Commit your changes and push the branch to your forked repository.
• Open a pull request on our repository.

Please make sure your changes are consistent with these common guidelines:

• Include unit tests when you contribute new features

• Keep API compatibility in mind when you change code

• Use messages as proposed here Conventional Commits when you commit your code, which refers to the Angular types:

  • build: Changes that affect the build system or external dependencies (example scopes: gulp, broccoli, npm)
  • ci: Changes to our CI configuration files and scripts (example scopes: Travis, Circle, BrowserStack, SauceLabs)
  • docs: Documentation only changes
  • feat: A new feature
  • fix: A bug fix
  • perf: A code change that improves performance
  • refactor: A code change that neither fixes a bug nor adds a feature
  • style: Changes that do not affect the meaning of the code (white-space, formatting, missing semi-colons, etc)
  • test: Adding missing tests or correcting existing tests

• Update the changelog: Change Log best practices

• We do use Semantic Versioning

• Please identify yourself with a valid email address as explained here: Setting your commit email address

• The content generated by the models in this repository is the sole responsibility of the user who initiates the generation. Users must ensure that their use of the generated content complies with all applicable laws, regulations, and ethical guidelines. The project contributors are not liable for any misuse, harm, or legal implications resulting from the use of the models provided. Users are encouraged to exercise caution and discretion when using generative AI tools.
• If you use the default background music, which is royalty free, you still must comply with Pixabay Content License Agreement (https://pixabay.com/fr/service/license-summary/).
• Too many parallel video generation processes might slow things down, just keep an eye on your local CPU usage. Happy coding! 🚀😊