AI-native animation studio · MONARKZ LLC
Original animated series, produced by an AI-native pipeline.
Zibotic creates family animated series with recurring characters — and the open-source production system that makes studio-style animation possible for a small, focused team, in many languages at once.
- Stage
- Pre-seed · Bootstrapped
- Markets
- US · UK · EU · Gulf
- Founded
- 2026
- Story & scriptLLM-assisted, human-directed01
- Character sheetLocked identity per character02
- KeyframesQwen-Image-Edit · Flux03
- AnimationWan 2.2 image-to-video04
- Voice & lip syncTTS · Wan S2V · Rhubarb05
- Multi-language releaseEN · AR · EU dubs06
The project
What Zibotic actually is.
Zibotic is an animation studio that makes its own cartoon series — short, funny, Pixar-style 3D stories starring a cast of characters that viewers meet again in every episode. Think of a family sitcom on a farm, told in one- to three-minute episodes and released several times a week on YouTube, TikTok and Facebook.
Traditional 3D animation needs a studio of modelers, riggers, animators and render farms, and months per episode. We are building a production pipeline that replaces most of that manual work with open-source AI models — while people keep control of what matters: the characters, the jokes, the story and the final cut.
Each character is designed once and locked as a reference sheet, so it looks the same in every shot and every episode. Scripts become shot lists, shots become keyframes, keyframes become animated clips, and an automated edit adds voices, music and subtitles. Because voices are separate from the animation, the same episode can be released in English, Arabic and European languages without animating it again.
The result we're working towards: original, consistent, studio-quality animated episodes, produced by a small team, at a cost that lets an independent studio publish every week.
What
Original animated comedy series with recurring characters
Where
YouTube (Shorts + episodes), TikTok, Facebook & Instagram Reels
Who for
Families and viewers 13+ in the US, UK, Europe and the Gulf
Languages
English first, then Gulf Arabic and European dubs
Formats
60–90 second Shorts and 2–3 minute episodes
How
Our own pipeline of open-source AI models on rented GPUs
Why now
Audiences are ready for independent animation.
Viewers now treat independent animated series as equals to studio shows — and they watch across languages. What has been missing is the ability to produce consistently, at quality, without a large studio budget.
of 14–24-year-old animation fans like independent YouTube series as much as, or more than, studio series
Source: YouTube, Animation Trends 2026of animation fans watch animated series in languages other than their own
Source: YouTube, Animation Trends 2026of children's searches in the movies / cartoons / TV category were for cartoons
Source: Kaspersky, Kids Report 2026related views for The Amazing Digital Circus — proof that independent animated IP can scale globally
Source: YouTube Culture & TrendsWhat we build
Two things, built together: the stories and the studio.
We are not an AI clip channel. We build long-lived characters audiences come back for, and the production system that keeps them consistent episode after episode.
Original animated IP
Recurring characters and worlds we own, written for family audiences in the US, UK, Europe and the Gulf. Story first, every episode.
An AI-native production pipeline
Open-source image and video models orchestrated end to end — from script to character sheets, keyframes, animation, voice and final edit.
Multi-language by design
One animation, many audiences. Narration and dialogue are produced per language — English, Gulf Arabic and European dubs — without re-animating.
Scope of work
From a script to a finished episode.
Heavy generation runs on GPU workers; everything else — writing, voices, timing, editing, localization — runs on CPU. Every stage is scripted, so episodes are reproducible and characters stay on-model.
Write
Series bible, episode scripts and shot lists with camera and emotion tags. Humans direct; AI accelerates drafts.
Design
Each character is designed once and locked as a multi-angle, multi-emotion reference sheet.
Stage
Keyframes for every shot combine locked characters with reusable 3D-styled locations.
Animate
Each keyframe becomes a 5-second clip with open video models on GPU workers, in parallel.
Voice
Character voices and narration per language, with automatic lip-sync timing.
Finish
Automated edit: merge, color-match, music, sound design, subtitles, 16:9 and 9:16 exports.
Consistency
Locked character sheets + keyframe-first animation keep every character on-model.
Efficiency
Short, independent GPU jobs run in parallel and retry automatically.
Reach
Audio-only localization: new languages without re-animating.
First series · in development
The Farmhouse
A reality-TV-style comedy on a farm where the animals live their dramas on camera — and the farmer just installed an over-eager AI assistant. Built for Shorts, Reels and TikTok, with longer episodes on YouTube.
Biscuit · Pug
Lazy, food-obsessed, secretly the smartest
Duchess · Pig
Farm influencer who thinks she's famous
Lorenzo · Llama
Dramatic diva who faints on cue
Bertha · Cow
The calm farm mom who runs everything
Ziggy · Zebra
Conspiracy theorist in a tinfoil hat
FARMI · AI speaker
Polite, far too literal, quietly taking over
Audience: 13+ and family viewing
Format: 60–90 s Shorts + longer YouTube episodes
Engagement: Audience comments shape upcoming episodes
Example episode
Episode 1: “The Farm Got an AI”
A 70-second pilot that shows the format: a topical premise, a reality-TV confessional style, fast jokes, a recurring cast and a cliffhanger.
0:00
Caption: "Day 1. The Farmer bought an AI." A glowing smart speaker — FARMI — switches on in the barn.
0:08
Confessionals: Duchess the pig is thrilled ("Finally, someone to film my content"). Doug the donkey is not.
0:16
FARMI "optimizes" Biscuit the pug from eleven meals a day to two. He freezes mid-chew.
0:24
Chaos: Woolly's sleep is cut, Lorenzo the llama is told crying is inefficient — and faints.
0:45
Bertha the cow calls a meeting. Biscuit asks FARMI the most efficient way to save power…
0:56
…and pulls the plug. The farm celebrates. Caption: "Day 2. The Farmer bought a BIGGER AI."
1:08
End card asks viewers what FARMI 2.0 should do next — comments help choose the next episode.
How this episode gets made
Script
Story, dialogue and a shot list with a camera angle and an emotion for every line.
16 shots
Each shot gets a keyframe with the locked characters, then a 5-second animated clip.
Voices
Ten distinct character voices, recorded per language, with lip-sync timing.
Final cut
Clips merged, color-matched, scored, subtitled and exported for every platform.
Roadmap
Where we're going.
Planned milestones. Dates are targets and will adapt to audience testing.
Pipeline & pilots
- End-to-end pipeline on GPU workers
- Character sheets for the first cast
- Pilot Shorts and first episodes
Season 1 launch
- Regular releases on YouTube, TikTok, Facebook
- Audience testing across pilot concepts
- English + Gulf Arabic
Scale & localize
- Multi-GPU / serverless production
- European language dubs
- Long-form episodes and compilations
Beyond video
- Licensing and brand partnerships
- Merchandise and books
- Pipeline tooling for other studios
Progress
What's working today — and what's next.
We're pre-launch. This is an honest snapshot of where the project stands.
Built and tested
- GPU environment running: ComfyUI with Wan 2.2 (image-to-video and audio-driven video models)
- CPU production server for scripts, voices, lip-sync timing and automated editing
- End-to-end test: script → voice → automatic lip sync → 3D render → subtitles → finished video
- Narration test: frame-by-frame scene analysis → timed voice-over → subtitled edit
- Automated thumbnail generation
- Series concept, 10-character cast and first episode script
In progress
- Pixar-style character design and locked reference sheets for the cast
- First fully animated pilot episode
- Multi-GPU job queue for parallel clip generation
- Channel launch and audience testing of pilot concepts
Business model
Characters first. Revenue follows the audience.
Our long-term value is the intellectual property — characters and worlds people recognize. Revenue streams unlock in stages as the audience grows. We are pre-revenue today.
Platform revenue
Ad revenue and creator programs on YouTube, Facebook and TikTok once eligible.
Brand sponsorships
Family-safe brand integrations and sponsored episodes as the audience grows.
Licensing
Licensing characters and episodes to platforms, apps and regional broadcasters.
Merchandise & books
Toys, stickers, apparel and picture books built on characters audiences already know.
Pipeline tooling
Offering our production pipeline to other independent studios and creators.
Compute & technology
Built on open models and elastic GPUs.
Our workload is GPU-intensive inference: thousands of short image-to-video and image-editing jobs per week, plus lightweight LoRA training. Cloud and GPU credits translate directly into more episodes, more languages and faster audience testing.
Models
- Wan 2.2 (I2V · S2V · 5B)
- Qwen-Image-Edit
- Flux
- Kokoro / ElevenLabs TTS
Tooling
- ComfyUI (API)
- Blender
- Rhubarb lip sync
- FFmpeg
- Python orchestration
Infrastructure
- RunPod GPU pods (RTX 5090)
- Azure CPU orchestration server
- Job queue across GPU workers
Workload profile
- Primary
- Inference — image editing, image-to-video, audio-driven video
- Secondary
- LoRA training for character consistency
- GPUs
- 24–32 GB (RTX 4090 / 5090); 48–80 GB for training
- Pattern
- ~1-minute jobs, parallel across 3–5 workers, serverless-friendly
- Crypto mining
- None
Responsible AI
Why this matters for the future.
AI can make animation accessible to independent creators everywhere — but only if it is used to create original work, not to copy it. These are the rules we produce by.
Original characters and stories only — no reuse of copyrighted footage or characters
Commercially licensed or CC0 assets, with credits where required
Clear AI-content disclosure on every platform
Family-safe content suitable for global audiences, including the Gulf region
Human creative direction and review on every episode
FAQ
Questions we get asked.
Is this just AI-generated content?
No. AI is our production tool, not our author. People write the stories, design and approve every character, direct each shot and review every episode before release. Every video is disclosed as AI-assisted, as platforms require.
How do the characters stay consistent?
Each character is designed once and locked as a multi-angle, multi-emotion reference sheet. Every shot starts from an approved keyframe built from that sheet, and automated checks reject shots where a character drifts off-model.
Do you use copyrighted material?
No. We only use our own characters and stories, open-source models with commercial licenses, and CC0 or commercially licensed assets — with credits where a license requires them.
Why multiple languages?
Half of animation fans watch series in languages other than their own. Because our voices are produced separately from the animation, a new language needs new audio — not a new animation.
What do you need from partners?
GPU and cloud compute is our main cost: every episode is many short image-to-video jobs. Compute credits and infrastructure partners let us produce more episodes, test more concepts and localize faster.
Where is the company based?
Zibotic is operated by MONARKZ LLC, registered in Farmington, Missouri, United States.
Let's build the future of independent animation.
We're open to startup programs, compute partners, co-production and early brand partnerships.
support@zibotic.comMONARKZ LLC
770 Market St
Farmington, MO 63640
United States