kunchenguid

kunchenguid / autopreso

Public

Realtime speech to presentation. Let the whiteboard whiteboard itself.

90
7
100% credibility
Found May 09, 2026 at 90 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
JavaScript
AI Summary

AutoPreso is an open-source web app that listens to your speech, transcribes it in real-time, and uses AI to draw and edit an Excalidraw whiteboard canvas automatically during presentations.

How It Works

1
🔍 Discover AutoPreso

You hear about a fun tool that turns your spoken ideas into drawings on a whiteboard as you talk.

2
🚀 Start the app

You easily launch it and your browser opens to a blank drawing canvas.

3
🎤 Pick your microphone

You choose your microphone so it can hear you clearly.

4
✏️ Sketch your starting scene

You draw a title, agenda, or simple diagram to guide the drawings.

5
🎯 Choose smart helpers

You select voice recognition and thinking helpers to match your needs.

6
▶️ Hit start and speak

You click the button, start talking, and the whiteboard comes alive with shapes and labels.

7
✨ Watch it draw live

As you explain ideas, it adds boxes, arrows, and notes perfectly timed to your words.

✅ Present hands-free

Your talk flows smoothly with visuals updating themselves, keeping everyone focused.

Sign up to see the full architecture

6 more

Sign Up Free

Star Growth

See how this repo grew from 90 to 90 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is autopreso?

autopreso turns spoken words into a live, evolving Excalidraw whiteboard via realtime speech to text. Run `npx autopreso` for a local web app: stage initial diagrams, pick your mic and realtime STT (Moonshine local or OpenAI realtime), then talk – transcripts stream to an AI agent (OpenAI, Codex, or Ollama) that draws shapes, adds labels, and rearranges in real time. JavaScript-powered, hands-free alternative to static slides for dynamic presentations.

Why is it gaining traction?

It hooks devs with true hands-free editing: speak concepts, watch structured diagrams emerge without mouse or keyboard. Local-first realtime speech recognition via Moonshine keeps it private and offline-capable, while model flexibility (bring your OpenAI key or Ollama) fits any stack. Excalidraw's infinite canvas plus viewport controls make visuals pop during talks, beating manual sketching.

Who should use this?

Tech presenters live-coding demos or architecture walkthroughs. Educators diagramming on the fly during lectures. Devs prototyping ideas in meetings who want realtime speech to presentation without prepping decks.

Verdict

Alpha with 90 stars and 1.0% credibility score means rough edges and breaking changes ahead, but quickstart and docs deliver a working prototype fast. Try it for macOS talks if you dig realtime STT github experiments – pairs well with Ollama for local runs.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.