AI Models and Service Availability
Current availability
VidRush AI Online is under development. The models listed below describe supported integrations and code defaults, not a claim that all are enabled in production. No live production model configuration has yet been verified for paid launch. Available engines and an exact quote must be shown before you authorize work. We do not provide an unrestricted generation service or imply endorsement by model vendors.
Scripts and planning
The text integration supports an OpenAI-compatible API and defaults to the specific model gpt-4.1-mini. A configured API endpoint or model may differ; its actual identity must be verified and disclosed before production activation.
Narration and transcription
The ElevenLabs narration integration defaults to eleven_multilingual_v2. The remote transcription integration defaults to OpenAI whisper-1. A local test path uses faster-whisper and a local Whisper model; the five-minute Apollo development sample used Whisper base.en for word timing and Kokoro af_heart for local preview narration. These test choices do not establish that the same voices or models are commercially available through this website.
Generated still images
The image-generation integration defaults to OpenAI gpt-image-1. Generated image availability depends on actual provider configuration and safety controls. No claim is made that a competitor’s advertised image model is used by VidRush AI Online.
Source-media matching and video assembly
OpenMontage supplies source-media discovery and CLIP-based matching in the development production pipeline. Source photographs and footage remain subject to their own licenses. Browser editing and video assembly use conventional media processing; assembled stock footage is not the output of a text-to-video model. The currently described workflow does not include a verified generative video model.
Safety and reporting
Input and output moderation must be configured and verified before paid AI production opens. This remains a launch requirement; this page does not claim that it is already complete. The Acceptable Use Policy explains prohibited material, human abuse-report review and appeals. Contact support@vidrush.dev with questions about the model or source used for a particular project.