Edição Automática com IA: como eu faço videos de Avatar IA em 20 minutos

Summary

Gabriel R. explains how he and his mentees use AI tools and automation to produce avatar-led YouTube videos quickly. His central business argument is that automation can lower production costs and increase output, but channel performance still depends on choosing strong ideas, angles, and titles—and producing content that feels credible and engaging.

Strategy and channel positioning

• AI-avatar channels can be built around many niches and sub-niches, including survival, mountain living, cooking, household tips, and gardening.
• A channel should offer a clear perspective or persona. The avatar and visuals need to look convincing enough to convey authority and earn audience trust.
• One growth tactic is to adapt successful content for other countries and languages. Gabriel describes translating or localizing titles and formats while moving quickly on emerging video trends.
• He argues that creators should focus less on making videos as cheaply as possible and more on producing work that meets a viable quality standard.

Production workflow and automation

The demonstrated workflow can automatically generate a script, narration, visuals, avatar segments, animations, and edits.

A typical setup involves:

1. Creating or importing a script and audio.
2. Adding an SRT subtitle file, if available, or generating one.
3. Selecting a visual style, directives, and character.
4. Choosing image or video models and animation formats.
5. Connecting an avatar account, selecting the avatar and background, and setting how often it appears.
6. Generating the video, then reviewing it and making any needed edits.

The avatar is used for short segments rather than throughout the whole video—roughly 10–15 cuts of 5–15 seconds each. The rest can use AI-generated images or video, or real footage.

Gabriel says the same automation can be run on multiple videos at once, freeing creators to spend more time on ideation, titles, new angles, and identifying what drives viral performance. He also notes that quality and workflow still need iteration. Planned improvements include AI-selected background music and sound effects timed to animations.

Metrics and examples cited

• Gabriel cites channels that reached roughly 200,000–500,000 views within 1–11 days. These are presented as examples, not typical or guaranteed outcomes.
• One mentee reportedly earned R$20,000 from AdSense and a further R$5,000 from products.
• Another mentee, Wellington, reportedly generated R$3,800, while two of his channels were not yet monetized.
• A 30-minute video reportedly cost about R$0.86 in one example, though Gabriel also estimates roughly R$3 for a finished video.
• A 30-minute video took 49 minutes to generate in moderate mode with a slow start. Gabriel says that setting can nearly double production time. Another project had been processing for 44 minutes.
• The avatar workflow reportedly used about 10–12 credits per video in one description; a sample with six cuts used 6–8 credits.
• The speaker claims that one automated video performed better than his earlier test but provides no specific performance metric.

Products, pricing, and business model

Gabriel promotes Radar Club, a course and trend-research tool. He cites an introductory price of R$7.90, followed by a recurring monthly fee to maintain access and updates.

He describes a separate automation tool with a cheapest plan of about R$100. A more advanced real-footage workflow was being tested in beta, with possible pricing of R$297–R$397; final pricing was undecided.

He frames paid tools as an investment in production capacity, while acknowledging that creators can build their own automation but may spend significant time developing and refining it. Revenue and pricing figures are promotional claims from the speaker and should not be treated as guaranteed results.

Practical recommendations

• Prioritize niche research, audience perspective, titles, and timely ideas; automation alone does not create a successful channel.
• Use the avatar selectively to manage credit costs, and combine it with other visual formats.
• Review generated videos before publishing, even when most of the process is automated.
• Evaluate tools by whether they support formats that work for the channel and provide dependable quality—not solely by lowest price.
• Treat cross-country adaptation as a speed-and-localization opportunity rather than assuming simple translation will guarantee traction.

Presenters and sources: Gabriel R. (presenter); examples and results attributed to his mentees and community. Tools and services mentioned include Radar Club, Invisible, HeyGen, GPT Image, Veo 3, and CapCut.