Automated video editing — GPT-4o · Whisper · ElevenLabs

Your editing style, replicated infinitely.

Upload a reference video, your raw footage and a topic. Mimics automatically edits your video in your style — script, voiceover and subtitles included.

No credit card — 3 free renders on signup

< 5 min

for a complete first render

4 steps

from raw footage to final video

50+ languages

supported by Whisper

Frame-perfect

subtitle synchronization

The problem

Video editing takes 4 to 8 hours per video for a solo creator.

You film, you reshoot the same lines ten times, you spend hours sorting through footage, writing scripts, syncing subtitles. And you start over the next week.

Existing automatic editing tools ignore your style. They output generic videos that don't look anything like what you do. The result is rarely publishable without spending as much time fixing it as editing from scratch.

Mimics starts from your editing style — not a universal template. It learns your pacing, structure and tone. And it replicates that on every new video.

The process

From raw footage to publishable video in 4 steps.

01

Reference style analysis

You upload a video whose editing you like — one of yours or another creator's. Mimics automatically extracts the average clip pace, cut frequency, vocal tone and narrative structure. This style profile becomes the base for everything that follows.

The model detects fast pacing (TikTok/Shorts), medium (YouTube 3-10 min) or slow (documentary/educational).

02

Script generation with GPT-4o

You describe your topic in a sentence. GPT-4o generates a spoken script calibrated to the analyzed style profile — word count, hook structure, sentence rhythm, tone (energetic, neutral or calm). The script is fully editable before render.

Word budget calculated automatically: about 2.5 words per second × target duration.

03

Smart footage selection

You upload your raw takes. Whisper transcribes every video word by word, identifies hesitations, duplicates and laughter. A scoring algorithm picks the best takes and orders them chronologically. You can force-include or exclude any clip in an interactive timeline before launching the render.

The last take of a repeated series is systematically favored. Hesitations and false starts are penalized in scoring.

04

Render, download and metadata

ElevenLabs generates the voiceover in your chosen tone or clones your voice from an audio sample. The video is edited, silences cut, subtitles synced to the frame. Once the render is done, you download the MP4 and generate optimized titles, descriptions and hashtags for YouTube or TikTok in one click.

The render continues server-side even if you close the page. Progress is automatically resumed when you return.

Features

Not an editor. A clone of how you think about editing.

Frame-perfect subtitle sync

Subtitles are generated on the original timeline via Whisper, then remapped after each silence cut using a two-pass TimeMap system. No drift, even on videos with 50+ cuts.

Repeated take detection and sorting

When you record the same line multiple times, Mimics identifies similar takes via text comparison and systematically picks the last one — your best version. Duplicates are removed without manual intervention.

Complete render isolation

Every render runs in an isolated UUID folder on the server. Even when multiple users start a render at the same time, their files never cross. The database applies Row-Level Security on every request.

Persistence after closing the page

The render runs on the server, not in your browser. If you close the page or refresh during processing, the app automatically detects the running render and resumes the progress display when you return.

SEO metadata generated automatically

Once the video is ready, GPT-4o generates 3 CTR-optimized title options, a full description and 10 relevant hashtags for the YouTube or TikTok algorithm — based on your script and style profile.

Voice clone or native audio

You can use an existing ElevenLabs voice, clone one from a 30-second audio sample, or use the original sound from your footage. Audio normalization is automatic — voice at -3 dBFS, ambient at -20 dBFS.

“Video editing takes 4 to 8 hours per video for a solo creator on average. Mimics cuts that to under 10 minutes.”

For a 60 to 90-second video, raw footage provided.

Who it's for

Built for creators who publish regularly.

YouTube creators

You publish 1 to 4 videos a month and you spend too much time editing. Mimics frees up 80% of your post-production time so you can focus on content and camera.

TikTok & Reels creators

Short-form requires fast pacing and a tight script. Mimics automatically adapts script density and cut frequency to TikTok/Shorts standards.

Agencies & studios

You manage multiple creators with different styles. Each style is saved as a reusable profile. One client, one profile. Zero retraining.

Pricing

Simple. No surprises.

Free

0€/ month

To test and validate.

  • 3 renders per month
  • 720p
  • AI script
  • Automatic subtitles
  • AI footage selection
Get started

Pro

29€/ month

For active creators.

  • Unlimited renders
  • 1080p
  • ElevenLabs voice clone
  • YouTube / TikTok metadata
  • 30-day history
  • Priority support
Start trial

Studio

79€/ month

For teams and agencies.

  • Everything in Pro
  • API access
  • Multi-workspace
  • Advanced analytics
  • 99.9% SLA
  • Dedicated onboarding
Contact us

FAQ

Frequently asked questions

What languages does Mimics support?

Mimics works with all languages supported by OpenAI's Whisper — French, English, Spanish, Arabic, and 50+ others. Scripts can be generated in any language you want.

Are my videos stored permanently?

No. Uploaded files are deleted automatically after rendering. Only the final MP4 is kept temporarily so you can download it. Each render runs in an isolated folder.

Can I use my own voice?

Yes. You can clone your voice from a 30-second audio clip via ElevenLabs, pick an existing voice from the library, or use your raw footage's original audio without a generated voiceover.

What happens if I close the page during render?

Nothing bad. The render keeps running on the server. When you come back to mimics.today, the app automatically detects the active render and resumes the progress display.

What's the maximum video length?

Maximum file size is 100 MB per file. For longer videos, we recommend splitting them into segments before upload.

Does Mimics generate subtitles automatically?

Yes. Whisper transcribes the voiceover with word-level timestamps. Subtitles are then remapped accounting for every cut to guarantee frame-perfect sync on the final video.

Start editing differently.

Create a free account