A good explanation deserves to be seen.

The hard part should be understanding an idea. Bringing it to life shouldn’t require learning an entire video editor.

LemmaReel connects the agent you already use to a local animation workflow. You bring the source material or a clear brief. Your agent shapes the explanation, and the CLI handles the visual templates, English narration, rendering, and technical checks.

Made for your material.

There is no bundled course or subject knowledge base. The example above uses a small synthetic dataset to demonstrate the animation workflow. Your videos start with your own files or ideas.

A tool you keep.

One purchase includes future updates to LemmaReel. Your installed version stays usable without periodic license checks. Your agent’s model access and usage are separate, and still require an internet connection.

What creates your video.

The visuals are programmatic animations rendered with Manim Community. English speech is synthesized locally with hexgrad/Kokoro-82M, using the kokoro-v1_0 model weights. LemmaReel does not provide a hosted image or video generation model.

Your chosen agent prepares the storyboard using the model configured in your own account. Codex, Claude Code and Gemini CLI are compatible agent interfaces, not a list of models supplied by LemmaReel. Model access, costs and provider policies remain with your chosen service. LemmaReel is an independent product and does not claim endorsement by those providers.

Our example video was planned with an AI agent and includes synthesized narration. Automatic checks verify technical properties such as layout, timing and media integrity; they do not certify factual accuracy or moderate the meaning of prompts and videos. See our acceptable use policy for content responsibilities.

Where we are today.

LemmaReel is in development. The local video pipeline has been exercised on an Apple Silicon Mac with Codex. Clean installation, Linux, and full Claude Code and Gemini CLI journeys are still being validated.