Building

ClipMesh

An AI studio for faceless videos: script, voiceover, b-roll and captions.

A desktop studio that turns an idea into a post-ready faceless video. AI writes the script and the voiceover, matches footage to every line and adds word-timed captions, and it all renders on your own PC.

Visit clipmesh.ai All builds

Status: Building
Category: AI video studio · Desktop app
Years: 2026

7
voice engines: three on the device, four in the cloud
31
caption style presets
12
language-model providers, including a local one
21
tools on the MCP server

The problem

Faceless channels run on volume: a script, a voiceover, matching footage, captions and an export for every video, on every channel. Doing that by hand is slow, and cloud tools add watermarks, render queues and monthly render limits.

What was built

A Tauri desktop app with a native Rust GPU renderer, AI script writing, text-to-speech voiceovers, automatic footage matching from the voiceover, semantic clip search and a multi-track timeline, backed by a cloud API for accounts, credits and payments.

Highlights

What it does best.

AI script writer
Hook-first scripts in 34 types, with the hook graded before anything is recorded.
AI voiceover
Lifelike voices in 30+ languages. No mic, no retakes.
AI footage director
Reads the voiceover and drops matching b-roll onto the timeline for every line.
Word-by-word captions
Transcribed on the device and frame-synced, in bold animated styles.
Renders on your PC
Your GPU does the work: no uploads, no render queue, unlimited exports.
Bring your own AI
Use credits, your own API keys, or local models through Ollama for free.