VisoTalk · Office #50,248

What it is

Calls itself “agent skill”.

• Calls itself: agent skill

“VisoTalk is an open-source agent skill for explainer videos.”

Where it puts itself

Its one line of difference: “Video tools pin every animation to a second, so change one sentence and everything after it lands a beat late. An explainer already has a timeline: its narration.”

“Video tools pin every animation to a second, so change one sentence and everything after it lands a beat late. An explainer already has a timeline: its narration.”

What it costs

No free plan.

• Free plan: no

“VisoTalk is early: the document format is 0.x and may change MIT licensed · © 2026 VisoTalk”

Who it sells to

Sells to developers.

• Sells to: developers

“Your coding agent writes the narration and the React that draws each part.”

How it runs

A web app. Runs locally and works offline. Open source. AI is the core of the product.

• Kind: web
• Runs where: on the device, offline
• Open source: yes
• AI role: core to the product

“Plays on the web, renders to video files.”

“Everything runs on your machine.”

What it works with

Plugin for Claude Code and Codex. Names Chrome, Liftoff and 2 more on its pages.

• Plugin for: Claude Code · Codex
• Names: Chrome · Liftoff · Why the Moon has phases · +1

“The first run sets up Chrome, the narration voice and the timing model”

“✓ skill added to Claude Code, Codex”

What it offers as proof

In beta. Try it without signing up. Publishes docs.

• Stage: beta
• Trial asks for: nothing

“VisoTalk is early: the document format is 0.x and may change”

“$ npx skills add visotalk/visotalk -g”