Practical overview
When it fits
Use it when you want a configurable prompt-to-video workflow and are prepared to supply the required local tools and third-party provider keys, or when evaluating the documented Colab and Premium API routes.
Who it is for
- Creators making short-form social video prototypes
- Developers exploring open-source AI video workflows
- Users producing educational or topic-driven videos from prompts
Documented uses
- Turn a topic or text prompt into a video concept and script
- Generate narration and align captions with speech
- Add AI-generated video segments and Pexels B-roll
- Render portrait or landscape compositions with configurable captions
Requirements & limits
The repository is public and identified with an MIT license. Documented access paths include Google Colab, local installation and a separate Premium API route. The evidence does not provide pricing or account requirements for those routes.
- Local installation requires Python 3.8+, FFmpeg and ImageMagick.
- Pexels is listed as always required in the configuration table.
- LLM, TTS and STT credentials depend on the providers selected.
- MuAPI is the documented video-generation backend connection.
- Portrait and landscape output use documented 1080x1920 and 1920x1080 settings.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Set up Text To Video AI locally
Prepare the documented local environment for prompt-based video generation.
Before you start
- Python 3.8 or later
- FFmpeg
- ImageMagick
- A local copy of the repository
- API keys for Pexels and any selected providers
Steps
Clone the repository and enter its directory.
Install the dependencies from requirements.txt.
Copy .env.example to .env.
Add the required API keys and select the LLM, TTS, STT and video settings in .env.
Expected outcome
A configured local project ready for the documented application command.
Guide limitations
- This guide covers setup only; the evidence does not establish execution beyond the documented command and output description.
- Provider-specific keys are conditional except for Pexels, which the configuration table lists as always required.
Configure a portrait short-form workflow
Set the documented output and caption options for vertical social video.
Before you start
- A configured project
- A selected video-generation and speech provider
- Caption settings in the .env file
Steps
Set VIDEO_ORIENTATION to portrait.
Enable captions with CAPTIONS_ENABLED=true.
Choose the documented caption font, color, face, stroke and position values.
Select the desired LLM, TTS and STT providers.
Expected outcome
Configuration targeting 1080x1920 portrait video with customizable captions.
Guide limitations
- The evidence documents configuration values but does not establish a completed render or platform upload.
- Portrait is described as recommended for Shorts, Reels and TikTok, not as a platform compatibility guarantee.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 3 October 2026
- Original website title
- SamurAIGPT/Text-To-Video-AI · Official homepage title · retained in its source language