NOTES IN, FILM OUT
Images, video & audioPraxis — notes in, a narrated film out
Give Praxis a topic, a PDF or your own notes: it drafts the lesson, draws the scenes, speaks the narration and records a captioned 1080p60 file — on your CPU.
Renders locally on your CPU. Nothing is uploaded.
Teaching is how you know you have learned — and Praxis turns that principle into output. Give it a topic, a PDF or the notes you already have, and it runs the whole production: the lesson gets drafted, the scenes get drawn, the narration gets spoken, and a captioned 1080p60 video file comes out the other end. The whole film studio lives in this tab.
The voices are the part that usually costs money, and they are all local: Kokoro, Piper with 900+ speakers, and Chatterbox for cloning — three neural voices that run on your machine and never touch a cloud API. The render, too, is yours: the CPU does the work, there is no queue to wait in and no credits to buy. You can render at 3 a.m. on a plane.
No AI provider configured? A deterministic offline planner still writes a complete, editable lesson, and the interface always labels which tier authored it. An offline draft can never masquerade as AI — you always know who wrote what, which is more honesty than most of this industry offers.
Use it to make revision videos from your notes, explainer clips from a chapter, or teaching material for the study group. Free, no account, nothing uploaded — the cast, crew and studio all live on your device.
AI-drafted lessons are labelled as AI; offline drafts as offline. Either way, the lesson is editable before it becomes a film.
What you can do
The parts people actually reach for.
The whole production
Draft, scenes, narration and render — one tool, one tab.
Three local voice engines
Kokoro, Piper with 900+ speakers, and Chatterbox for cloning.
CPU render, no queue
1080p60 encoded on your machine — no credits, no waiting.
Captioned output
Videos come out with captions baked in.
From your material
A topic, a PDF, or the notes you already wrote.
Free and offline-capable
No account, no upload, no cloud APIs anywhere in the loop.
How it works
Three steps, no detours.
Choose the source
A topic, a PDF, or notes from the workspace.
Draft and voice it
The lesson is written, the scenes drawn, the narration spoken — locally.
Render and download
A captioned 1080p60 file, encoded on your CPU.
Questions, answered
The ones people ask first.
Does video rendering need a subscription?
No. Rendering and narration run on your CPU — no credits, no queue, no account.
Which voices can narrate?
Three local neural engines: Kokoro, Piper (900+ speakers) and Chatterbox for cloning. None is a cloud API.
What if I have no AI provider set?
A deterministic offline planner still drafts a complete, editable lesson — always labeled as offline.
More in Drazil
One workspace, 44 tools — all free, all local.
Renders locally on your CPU. Nothing is uploaded. · or browse all 44 tools