Ready
Queue
Ready
0 credits

AI Video Studio
Locally Powered AI

Generate videos, talking avatars, images, voiceovers, and audiobooks — all running locally on your hardware. 5+ video engines, 6 lip sync engines, 5 TTS voices, zero cloud dependency.

🎬 5+ Video Models
🗣️ 6 Lip Sync Engines
🔊 5 TTS Engines
📖 6 Export Formats

Content Creation

Generate video, images, speech, and music with AI — from a single prompt to a finished production.

🎬

AI Video Generation

Multiple AI engines for image-to-video and text-to-video with draft and premium quality presets.

Create Video →
🖼️

AI Image Generation

Text-to-image, img2img, inpainting, background removal, watermark removal, and AI face enhancement.

Generate Image →
🗣️

Talking Avatars & Lip Sync

6 lip sync engines with automatic cascade fallback for realistic talking head videos.

Create Avatar →
🔊

Text-to-Speech

5 text-to-speech engines with 20+ voices and voice cloning support.

Generate Speech →
🧠

AI Narrative Engine

AI-powered story parsing with automatic scene splitting and vision analysis for context-aware generation.

Write Narrative →
🎵

Video Stitching & Music

Multi-segment concatenation, audio normalization, and background music mixing for polished final output.

Browse Music →

Publishing & Documents

From manuscript to finished book or marketing campaign — write, compile, and publish.

📖

Book Studio

Upload PDF, DOCX, EPUB, TXT, RTF, or HTML. Scrivener-style binder, Quill rich editor, audiobook generation, and compile to 6 export formats.

Open Studio →
📊

UGC Video Templates

9 ready-made templates — Hands Holding, Unboxing, Bedroom Review, Car Review, Lifestyle, Studio Shot, Close-up, Rotation, Comparison.

Browse Templates →

Developer Tools

A full IDE and training dashboard built into the platform.

💻

Code Editor (IDE)

Full IDE with 4-panel layout, AI coding agent, file explorer, integrated terminal, and diff preview.

Open Editor →
🧪

Training Dashboard

Real-time ML metrics with live charts, fine-tuning pipelines, auto-recovery, and AI-generated training reports.

Open Dashboard →

Platform

⚡

Hardware Accelerated

GPU-accelerated generation with BF16/TF32 precision and torch.compile acceleration.

🔒

100% Local & Private

Powered by Ollama + ComfyUI. No cloud services, no data leaves your machine, no subscription required. Full control over your AI stack.

5+ Video Models
6 Lip Sync Engines
5 TTS Engines
9 UGC Templates
Local GPU Compute

The AI video generator built for video creators

📷Drop image, click to upload, or paste from clipboard

Describe the motion, camera movement, or action you want to see. The AI will animate your image based on this description.

seconds into video
The character will speak this text with lip-synced animation.
Checking memory...

Context Preview — edit AI-generated per-segment prompts before generating

Generation History

▼

The AI image generator for video creators

Text to Image Image Edit Enhance Photo Watermark Removal
📐

AI Voiceover

Create talking avatars or generate audio from text

1. Upload Avatar Image

🎭

Drop face image here or click to browse

Best results: Clear front-facing photo

2. Enter Text to Speak

3. Voice Settings

Default engine: Good quality, runs locally, no watermarks

Tips for Best Results

  • Use a clear, front-facing photo
  • Good lighting improves lip sync
  • Keep text under 500 characters
  • Default engine works best for most cases

Text to Speech

Generate natural-sounding voiceovers with AI

Enter Your Text

0 characters

Voice Settings

Audio Result

Tips

  • VibeVoice offers the most natural sound
  • Use punctuation for natural pauses
  • Preview voice before generating
  • Edge TTS has more voice options

Enhance Photo

Upscale images using AI enhancement

🖼️

Click or drag images here

You can select multiple images

2x is faster, 4x provides higher resolution output.

Creative Assets

Videos

Click to play. Select multiple to delete.

Edited Photos

Photos edited with ChronoEdit. Select multiple to delete.

Generated Images

AI-generated images from Text-to-Image. Select multiple to delete.

Merge Videos

Drag videos to reorder. They will be seamlessly stitched together.

🎬

Click or drag videos here

Filters

⌘K

Filters

⌘K

AI Prompt Enhancer

Enhance your prompts with context-aware AI

Enhance Your Prompt

Project Context Optional

e.g., CLAUDE.md, README.md

OR

Recent Enhancements

No enhancement history yet

Marketing

GPU Memory:
Loading...
🤖

UGC Video Templates

Select a template style for your product video

Loading templates...

Recent Generations

LoRA Training

Idle
Started --
Elapsed --
ETA --
▼ Training Content
Loading training data info...
0%
System Memory 0 GB
Loading...
GPU Utilization 0%
0 F
Training Loss --
Learning Rate --
Progress 0/0 ep
0%
Phase0/8
Epoch0/0
Step0/0
Training Speed -- step/s
No reports yet
Idle
📓

Select or create a notebook to get started.

Phone Agent

Paid APIs used: no SIP/PSTN ready, not configured
BackendLoading
Active Calls0
Calls Today0
Leads0
Appointments0
Errors0

Runtime Status

Recent Calls

No calls loaded.

Browser Test Mode

No active call
Start a browser test call to begin.
STT status loads from runtime discovery.
TTS fallback status loads from runtime discovery.
Latency: not measured yet.

Transcript

Voicemail

Appointment Requests

Agent Identity

Leads

Create / Edit Appointment

Appointments

Call History

Discovered Providers

Telephony

SIP/PSTN ready, not configured
Browser Test ModeAvailable in v1
PSTN/SIPNot configured
AsteriskNot detected

Asterisk pjsip.conf template

Load telephony config to generate template.

Asterisk extensions.conf template

Load telephony config to generate template.

Podcast

Scheduler Loading...

Generate

Create a new podcast from your audiobook library, narrated in your trained voice. Daily scheduled run targets 15 minutes; manual runs can be any length.

Idle
Uses ~165 words/minute to calculate the maximum script length.
If set, the script must land near this target. Cannot exceed max time.
Default: 2250-2750 words (~15 minutes).

History

All previously generated podcasts. Download individual episodes or batch them as a ZIP.

Title Generated Topics Duration Size Status Actions

Loading podcast history...

Web

Public webpage extraction, media downloads, and full-site mirrors. Only download content you own or have permission to archive. Some platforms may restrict automated downloads.

⚖️ This tool does not bypass logins, paywalls, CAPTCHAs, or other access controls. Respect robots.txt, platform terms, and copyright.

Smart Scraper

Analyse a public page and extract structured fields.

Custom selectors

Pagination

Export format

Preview

Run Analyze Page to preview detected content.

Media Downloader

Image & video extraction from a public page.

Live progress

Files found0
Downloaded0
Skipped0
Failed0
Estimated ZIP size0 B

Full Site Downloader

SiteSucker-style mirror of a public site for offline browsing.

Crawl scope
Include asset types

Crawl report

Pages crawled0
Assets downloaded0
Errors0
Total downloaded0 B

Job history

Name URL Mode Status Found Downloaded Created Size Actions
No jobs yet.

Scraper settings

Book Writer

Write books chapter-by-chapter or generate professional audiobooks with AI voices.

Binder

No chapters yet.

0 chapters
📝

Select a chapter from the Binder to begin writing.

Inspector
Table of Contents
Metadata
Status -- Words 0 Last saved --
Reading Voice
Chapter Notes
Generation
Guardrails
Source Content
TXT, PDF, DOCX…
Style Sheet

Sermon → Paperback Book

Upload a sermon or lecture audio file. Taleclip transcribes it, then converts the transcript into a chaptered paperback manuscript with strict editorial guardrails (no fabrication, preserves the speaker's voice, no lecture artifacts). Export to Markdown, plain text, or 6×9 KDP-ready DOCX.

1. Upload audio 2. Transcribe 3. Review transcript 4. Generate book 5. Export
🎙️

Drop audio/video here, or click to browse

.mp4, .mkv, .mov, .webm, .avi, .mp3, .wav, .flac, .ogg, .m4a, .aac (max 500 MB)

Already have a transcript? Paste text instead

Print Cover Calculator

Calculate cover dimensions and generate print-ready templates for KDP paperback and hardcover books.

📄
Drop .docx or .pdf to auto-detect page count
24–830

Cover Image Compositing ▼

Image Enhancement ▼

Template Preview

📖
Enter book details and click Calculate to see template preview

Manuscript

0 words 0 chars
?
PDF, DOCX, EPUB, TXT, RTF, HTML

Voice

Output

Live Chunks

0 / 0
Chunks will appear here as they generate.

Audio Files

No audiobooks generated yet.

Voice Training

How Voice Training Works

Train the AI to speak in your voice for audiobook narration. There are two approaches — pick whichever fits your situation:

QUICK START (Zero-Shot Cloning)

Upload or record a voice sample (10+ seconds). The AI imitates your voice immediately — no training wait. Good enough for drafts. Uses Chatterbox engine.

BEST QUALITY (XTTS Fine-Tuning)

Upload 60+ seconds of audio, then train a custom model on your voice (takes 1–2 hours). The AI truly learns your voice — much more accurate and natural. Best for final audiobooks.

1

Upload .wav or .mp3 audio files of you speaking clearly. You can also click Record to record directly from your microphone. Tips for best results:

  • Use a quiet room — no background noise, music, or echo
  • Speak naturally at your normal pace and tone
  • For Quick Start: 10–30 seconds is enough
  • For XTTS Fine-Tuning: upload at least 60 seconds total (more is better, 2–5 minutes ideal)
  • Multiple short clips are fine — they get combined automatically
2a

Click the button below and you'll be shown 5 short passages to read aloud. The system records you reading each one, then builds a voice profile from those recordings. This is the fastest way to get a working voice clone.

After completing this, your voice will appear in the Voice dropdown on the Audiobook tab (Chatterbox engine).

2b

Similar to Guided Training, but uses excerpts from your actual book as the reading material. Each round has 3 passages. The more rounds you do, the better the voice quality gets. This also adds to your voice sample data for XTTS training below.

3

This trains a dedicated AI model on your voice data. Unlike zero-shot cloning (Steps 2a/2b), this actually learns the unique characteristics of your voice — pitch, cadence, tone, pronunciation. The result is a much more accurate and natural-sounding voice.

How to use XTTS Fine-Tuning:

  1. Upload voice samples in Step 1 above (minimum 60 seconds total, 2–5 minutes recommended)
  2. Click "Prepare Data" — this splits your audio into clean training chunks, normalizes volume, and creates a training dataset. Wait for it to finish.
  3. Click "Train My Voice" — this starts the actual AI training. It runs for the number of epochs shown below (default 50). Training takes 1–2 hours depending on data size. You can leave this page and come back — training continues in the background.
  4. When training completes, click "Test Voice" to hear a sample of the AI speaking in your trained voice.
  5. Go to the Audiobook tab, select "XTTS (Fine-tuned)" from the Voice dropdown, and generate your audiobook.
Status Checking...
Advanced Settings

Visual Education

Convert documents into narrated educational videos

Source Documents

Drag & drop documents here, or click to browse

Supports .docx, .pdf, .txt, .pptx

— OR browse server folder —

    Settings

    Claude CLI analyzes documents deeper and produces richer transcripts
    2 min
    ▶ Pronunciation Guide — fix how TTS says specific words