CLONER une vidéo avec l'IA en 5 minutes (outil caché de GEMINI)

CLONER une vidéo avec l'IA en 5 minutes (outil caché de GEMINI)

CLONE a video with AI in 5 minutes (hidden GEMINI tool)

Title history
  • CLONER une vidéo avec l'IA en 5 minutes (outil caché de GEMINI) — observed on 20261112-0057 UTC (YouTube)

by Thomas Berton · View the channel on PSI

Applied Sciences & Engineering AI — Models & Research UYQArtificial intelligence (AI)UYQFGenerative artificial intelligence
🎙 Thomas Berton 👥 48K 📅 February 12, 2026 ⏱ 21 min 👁 7K 📄 tutorial 🧭 2026-11-12
Available in: English (current) Français

Keywords

video cloninggenerative AIno-code workflowAirtableadvertising

Summary

The video presents a method to clone existing videos using AI, specifically leveraging Google’s hidden Gemini video analysis feature. The creator, Thomas Berton, demonstrates a complete workflow using N8N for automation and Airtable for data management. He explains that the system analyzes a video to extract its style, transitions, colorimetry, and effects, then regenerates a new version with different characters or styles. The tools involved include Nano Banana Pro for image generation, Kling 2.6 Pro for video generation, and Suno for music, all orchestrated via an AI agent. The tutorial covers accessing AI Studio for video analysis, setting up N8N workbooks, and using WaveSpeed AI and Kai.ai as model aggregators. Business use cases for e-commerce and agencies are discussed, and the creator promotes his paid club for templates. He also shows how to download Instagram videos using a Chrome extension. The video ends with a quick guide to creating motion control videos with Cling 2.6.

155 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video is highly valuable as a practical, hands-on guide. The creator demonstrates real results, showing a side-by-side comparison of an original Prada ad and his cloned version, which convincingly illustrates the effectiveness of the workflow. The argumentation is based on direct demonstrations and cost comparisons, claiming a production cost of €4 versus an estimated budget of €5,100 for the original. The explanation is well-structured, breaking down each step and tool. However, the argumentation relies solely on the creator’s own experience and does not provide external validation or compare with alternative methods, which limits its objectivity.

Scientific Rigor, Source Quality, Title Accuracy

The video does not cite external scientific sources; it primarily references the creator’s own resources and promotional links to tools like AI Studio, WaveSpeed AI, and n8n. The claim that the ‘hidden Gemini tool’ is overlooked is supported only by the creator’s assertion. The title perfectly matches the content, and the step-by-step process is clear and reproducible. The video is more of a tutorial than a scientific analysis, so the lack of peer-reviewed references is expected, but the promotional tone for the paid club slightly reduces its credibility. The inclusion of a sponsorship segment is visible but not mentioned in the description, and the video does not address potential biases.

221 words

Title / Content Match

The title accurately describes the content: the video demonstrates how to clone a video using AI in about 5 minutes, and the hidden Gemini tool is indeed presented.

Quality & Reliability

7/10

The video provides a thorough step-by-step tutorial with practical demonstrations, but relies on the creator's personal experience and promotional links, lacking independent verification or scientific sources.

Chapters

Cited Sources

  • AI Studio Google — Platform used to access Gemini models and analyze videos.
  • WaveSpeed AI — Model aggregator used for image and video generation via API.
  • n8n — No-code automation platform used to orchestrate the workflow.
  • AIOS.fr — Creator's service for AI projects.
  • LinkedIn - Thomas Berton — Creator's professional profile.

External References

Contribution & Novelties

The video’s originality lies in combining Gemini’s video analysis capability with a no-code automation pipeline (N8N + Airtable) to replicate the style of any video. It provides a practical workflow that uses distinct AI models for images, video, and music, orchestrated via a single system. The SILCAM prompt framework is a notable adaptation for structured video analysis and prompt generation.

60 words

Radar Profile

The radar profile shows high scores in information quantity (8) and technical level (8), indicating a dense, practical tutorial. The lower scores in information quality (7) and reliability (6) reflect the lack of external validation and reliance on personal experience, making it a useful but not academically rigorous resource.

Reliability 6/10