In the making · September 16, 2026

Keep the medium. Shape the thought.

A future direction for refining audio and video while keeping the result in the medium you chose.

MeerKatta begins with a simple idea: capture the thought before the effort of editing interrupts it. Today, that means turning rough speech into useful text. But text is not always the finished form. A podcast thought may belong in audio. A demonstration or story may need to remain video.

From recording to a clearer recording.

We are exploring a workflow where you can speak or record freely, then ask MeerKatta to help shape what you made. It could remove filler and repetition, tighten pauses, or restructure sections so the idea is easier to follow—while preserving your meaning and your voice.

The medium would stay the same. Audio would become refined audio. Video would become refined video. The goal is not format conversion. It is to give a recorded thought the same path from rough capture to useful result that spoken words have today.

AI guides the edit. Processing tools do the work.

In this direction, AI acts as the editorial brain: it helps understand the request, recognize the structure, and decide what should change. Local processing and specialized audio or video services would carry out the work, such as making cuts, repairing speech, or synchronizing revised words with video when the intent is clear.

Changes that could affect meaning, voice, or appearance should remain visible and under your control. The useful result is still yours—not a generic replacement for what you meant to say.

For podcasts, videos, and thoughts worth keeping.

You might use it to clean a spoken reflection before sharing it as a podcast, tighten a product explanation for YouTube, or keep a personal recording that is easier to return to. Different channels need different processing, but the underlying practice stays consistent: capture first, preserve the intent, then shape the result.

In the making, not in the current release.

Audio and video refinement are directions we are exploring, not features available today. The workflow, service choices, privacy model, quality bar, and release timing are not settled. We are sharing the direction early because the useful moments should shape what gets built.

Would you want to refine audio, video, or both? You can vote on the homepage or tell us what you would record and where you would share or keep it.