VidNexusVidNexus
v1.0.0·Now shipping
Early Bird // 10/10 seats·$999 lifetime

One prompt in.
YouTube-ready video out.

In under 10 minutes.

VidNexus is the local-first AI studio for faceless YouTube creators. Script, voiceover, stock footage, motion graphics, final render, YouTube publishing - long-form or Shorts, every step connected, all on your hardware. No cloud, no per-render fees, no privacy headaches.

Windows 10/11 today · Mac & Linux on roadmap · 14-day refund

VidNexus in 90 seconds
▶ play
[01]GPU-accelerated · NVIDIA NVENC + AMD AMF//
[02]Local-first · your data stays on your machine//
[03]100+ motion graphic types · auto-suggested//
[04]16:9 long-form + 9:16 Shorts//
[05]Publish straight to YouTube//
[06]8-min video · built in about 10 min//
[07]Zero per-render fees · pay once, render forever//
[08]14-day money-back guarantee//
[09]GPU-accelerated · NVIDIA NVENC + AMD AMF//
[10]Local-first · your data stays on your machine//
[11]100+ motion graphic types · auto-suggested//
[12]16:9 long-form + 9:16 Shorts//
[13]Publish straight to YouTube//
[14]8-min video · built in about 10 min//
[15]Zero per-render fees · pay once, render forever//
[16]14-day money-back guarantee//
Connected to the ecosystem

Bring your own API keys for the services you already use.Zero markup. Zero lock-in.

ElevenLabs
Voice synthesis
ai33.pro
More voices, for less
Pexels
Stock footage
Pixabay
Stock footage
Coverr
Stock footage
Wikimedia
Open media
Higgsfield
Cinematic AI shots
Ollama
Local LLM
OpenAI
Hosted LLM
Claude
AI editor
Pikzels
AI thumbnails
Mapbox
Map graphics
YouTube
Publishing + CC search

Plus FFmpeg with NVENC/AMF, Whisper for captions, Remotion for motion graphics, SQLite for local data, and any OpenAI-compatible endpoint you point at (OpenRouter, Together, LM Studio, vLLM).

// workflow.exec

Three stages.
All connected.

Every stage is reviewable. You approve the outline before the script, and the script before voiceover. You approve scenes before motion graphics. You stay the editor - VidNexus connects the pipeline.

pipeline.integrity
status: nominal
prompt
review
render
STAGE.01
ready
01
ready

Prompt your idea.

Type a topic, paste your own script, or upload your own voiceover. The LLM drafts a researched outline you review and edit, then streams the full script section by section - with a hook score out of 30 and a fact-check panel that suggests rewrites.

New project · Step 1 of 3
Topic
Make a documentary about food preservation before refrigerators. Focus on root cellars, salt curing, and ice harvesting.
Documentary
10 min
History Kit
STAGE.02
active
02
active

Review every scene.

Video settings surfaces every override - brand kit, voice, speed, 16:9 or 9:16 Shorts format, language, compliance sources - with a cost estimate before you build. Regenerate a single scene without re-running the whole pipeline.

Video settings · Scene 04 of 12
Voice
Adam · 1.0×
Theme
Vintageoverride
Stock
Pexels, Coverr
Motion
8 suggested
Music
Ambient · 78 BPMoverride
STAGE.03
queued
03
queued

Render on your GPU.

Final MP4 via NVENC on NVIDIA, AMF on AMD. A 5-minute video takes about 6 minutes to build, with voice, footage and graphics included. Publish straight to YouTube, or export FCP XML for Premiere/Resolve and SRT subtitles in the same pass.

Building · 03:40 elapsed
✓
Script analyzed
12 scenes
✓
Voiceover
12/12 · ×4
✓
Motion graphics
6 rendered
•
Footage search
9/12 · ~30s
○
Final render
NVENC · 45s
Estimated total
6m 05s
total runtime // about 10 to 12 min from topic to finished video · about 6 min of it is the unattended build
Built for the format

Any faceless niche, end-to-end.

The motion-graphic library, music sync, and pacing analyzer adapt to your content type. Whether you ship a daily Top 10 or a weekly hour-long documentary, the pipeline fits.

U01 · Documentary
supported

Long-form narrative builds with multi-scene b-roll cycling, beat-snapped music, and motion-graphic infographics.

U02 · Top 10 / Listicle
supported

Per-item title cards, automatic countdown counter, headline transitions between sections.

U03 · Educational explainer
supported

Charts, comparisons, decision flowcharts, and bullet-list overlays - all auto-suggested by the LLM.

U04 · History / culture
supported

Map animations, timeline scrubs, quote cards, and equipment schematics for context-heavy storytelling.

The team you don't need to hire

Eight roles.
One nexus.

A full faceless YouTube production crew costs $0 per video at freelance rates. VidNexus connects all of it for $1,499 once - and you stay the creative director.

What VidNexus automates
scope

The pipeline. Not the creative direction. You still pick the topic, set the tone, set up the brand kit, and approve every scene before render. VidNexus handles the assembly - research, drafting, voicing, sourcing, animating, cutting, captioning, encoding, publishing.

ps · running roles
8 processes
PID
Role
Cost/run
1000Researcher[killed]~$200
1007Scriptwriter[killed]~$300
1014Voice actor[killed]~$150
1021Stock footage curator[killed]~$100
1028Motion designer[killed]~$400
1035Video editor[killed]~$250
1042Captioner[killed]~$60
1049Thumbnail designer[killed]~$80
0001vidnexus.exe[running]$1,499 once
System metrics
live · v1.0.0
M01
~0
Min build · 5-min video
M02
$0
Per-render fees
M03
0+
Motion graphic types
M04
$0
One license · forever
Cloud SaaS vs. VidNexus

Cloud tools tax you per render.
VidNexus doesn't.

Diff · saas // vidnexus
6 rows
Feature
Cloud SaaS
VidNexus
Pricing model
$30–100/mo subscription, forever
$1,499 one-time. Forever.
Per-render fees
$1–5 per minute of output
$0 per minute, always
Where your data lives
Their servers, their access
Your machine, your access
Render speed
Cloud queue + upload/download
GPU-direct on your hardware
Offline use
Internet required, always
Optional - cache + run offline
AI model choice
Locked to their model
Any OpenAI-compatible LLM
The pipeline

Every step. On your hardware.

P01 · AI script generation
shipped

Any OpenAI-compatible LLM. Ollama locally, OpenAI in the cloud - your call.

P02 · Multi-source stock footage
shipped

Pexels, Pixabay, Coverr, public-domain archives (Internet Archive, Wikimedia Commons, NASA, DVIDS) and your own downloaded clips. CLIP-scored against the narration, with the right moment picked from long archive films.

P03 · AI voiceover
shipped

ElevenLabs or ai33.pro voices, or your own. Per-scene caching, parallel synthesis, a voice per channel.

P04 · 100+ motion graphic types
shipped

Title cards, charts, counters, maps, comparisons, callouts - drag-and-drop or AI auto-fill, all editable.

P05 · Styled captions
shipped

Whisper transcription per scene, caption styles, burn-in optional, SRT export.

P06 · GPU-accelerated render
shipped

NVENC on NVIDIA, AMF on AMD. ~5–10× faster than CPU encoding.

P07 · AI thumbnails
shipped

1-8 variants per run via Pikzels, prompt suggested from your script. Pick the winner.

P08 · Multi-track timeline editor
shipped

Real footage thumbnails, waveform silhouettes, motion-ribbon overlays.

P09 · Privacy by design
shipped

No content uploaded. Disconnect the internet - renders still work.

P10 · Long-form + Shorts
shipped

16:9 from ~1 min to 25-30 min deep dives, or 9:16 Shorts for YouTube Shorts, TikTok and Reels.

P11 · YouTube publishing
shipped

Upload with title, description, tags, thumbnail, playlist and schedule. Private by default.

P12 · Research toolkit
shipped

Topic pulse, keywords, competitor teardown, scored titles, tags and auto-chapter descriptions.

// initiate connection

Connect the pipeline.

First 10 customers pay $999 lifetime. After that it's $1,499. Either way: one license, forever, on your hardware. Waitlist members get a 24-hour heads-up before Early Bird opens.

vidnexus install --license=lifetime --tier=early-bird

One email at launch · No spam · No drip campaign