Recora Screen.
Star on GitHub ↗Get the app ↗

OPEN-SOURCE SCREEN RECORDER FOR MACOS

Record once. Edit less. Edit with AI. Share faster.

CodexClaudeWORKS WITH CLAUDE & CODEX · MCPDirect the whole studio.Create projects, edit timelines and export by asking.

The open-source screen recorder you can control with Claude and Codex.

Record your screen, camera and audio, or import a video and start shaping it.
Automatic zooms follow the action. Every zoom, layer and camera moment stays editable. Connect Claude or Codex through MCP, transcribe locally, then export up to 4K.

Download for macOS ↓Connect Claude + Codex ↓

Apple silicon · macOS 15+ · Local projects · MIT licensed

Recora Screen editor with a video filmstrip, editable zoom and timed text layer
The current editor. Original media stays untouched while clips, zooms and layers remain editable on the timeline.
YOUR WORK. YOUR DEVICE.

No upload before you can edit.
No account between you and your next idea.

LOCAL
BY DEFAULT ↗

THE PRODUCT, TODAY

A focused studio.
From first take to final file.

Start in a local project library, make precise changes on a visual timeline, and keep every effect available for another pass.

01 / LOCAL WORKSPACE

Your recordings have a home.

Record or import, reopen any project, and carry the whole project folder with you. No account is required to organize or edit your work.

Recora Screen local project library with one clean demonstration project
02 / EDITABLE FOCUS

Zooms you can still change.

Clicks, drags and typing can become focus moments. Adjust timing, depth, motion and cursor following without baking the effect into your recording.

Zoom and cursor controls beside an editable zoom track
03 / TIMED LAYERS

Show it only when it matters.

Add text, images, arrows, blur or solid covers. Move layers on the timeline and drag their edges to control exactly when they appear.

Layer controls and a timed text layer in the Recora Screen timeline

FROM FIRST TAKE TO FINAL CUT

Everything your story needs.
Room to make it your own.

01 / CAPTURE

A screen. A camera.
Your point of view.

Record a screen, window or region with separate camera, microphone and system audio tracks. Mirror your camera, tune its corners and shadow, or let it shrink automatically during zooms.

02 / EDIT

Find your flow.

Select a clip to see its controls. Filmstrips and audio waveforms help you find the moment; separate Zoom and Layers rows keep effects visible. Split, delete, trim or reorder, with actions that name what they change. Undo keeps the original media safe.

03 / FOCUS

Point out what matters.

Continuous typing holds the zoom steady. Edit zoom settings in a dialog without losing your list position. Change a selected camera clip’s appearance or the video default, with the scope shown beside the controls. Add arrows, blur and covers.

04 / A LITTLE HELP

Less busywork.
More of your voice.

Transcribe Turkish, English and more with local Whisper. Edit through your words and review suggested cuts from your microphone pauses. Silence cleanup requires a recorded microphone track. Audible system audio is protected by default. Bring your OpenAI or Anthropic key for optional editing assistance and YouTube copy.

00:12 Let’s start with the idea.
00:16Then make it our own.Whisper runs locally after model download.
05 / FRAME

Give the story room.

Choose solid, gradient or image backgrounds. Set the aspect ratio, padding, rounded corners, shadow and browser frame for every destination.

06 / KEEP CREATING

Every idea gets a home.

A focused branded opening takes you straight into your local workspace. Start with Record screen or Import video, follow the first-edit guide, and reopen Quick start whenever needed. Automatic saves and portable project folders keep your work ready for the next session.

07 / SHARE

Ready for your audience.

Export H.264/AAC MP4 up to 4K or a looping GIF up to 60 seconds. Burn in captions or save SRT/VTT. Your project remains editable after export.

AI, WITH A CLEAR JOB

Ask for the edit.
Keep the control.

Use the optional in-app assistant with your own OpenAI or Anthropic key, or connect Claude and Codex through the bundled MCP server. The same validated editing commands power both the visual editor and agent workflows.

  • Transcribe on your Mac with local Whisper.
  • Review suggested microphone silence cuts before applying them.
  • Create projects, edit timelines, inspect previews and export through MCP.
  • Undo agent-made timeline changes in the app.
See the MCP workflow ↓
Recora Screen AI assistant panel beside the visual timeline
AI is optional. Local recording and editing do not require an API key.

ONE IDEA. THREE SIMPLE STEPS.

Less setup.
More showing.

  1. 01

    Record or import.

    Record screen opens capture setup. Import video opens an existing MP4, MOV or M4V.

  2. 02

    Keep the good bits.

    Select a clip, split it and remove an unwanted part. Adjust camera or zoom where needed. Your edits save automatically.

  3. 03

    Put it out there.

    Export your video. Share your tutorial, walkthrough or next big idea.

YOUR ASSISTANT. YOUR STUDIO.

Say it.
Shape it.

Connect Codex, Claude Code or Claude Desktop to the bundled local MCP server. Create projects, control recordings, edit timelines and export through the same validated commands as the app.

Read the MCP setup ↗
Codex ↗Claude Code ↗Claude Desktop ↗
RECORDED WITH CODEX · ACTUAL MCP RESULT

Add a 1.6× zoom, place a timed title, and export a 1080p MP4.

ONE EDIT, FROM PROMPT TO EXPORT

timeline_apply → preview_frame → export_start

Actual app export from a clean 12-second fixture; the final frame holds for 3 seconds. Captions explain the recorded tool calls. No camera or audio is used in this sample.

The bundled server exposes the same project, timeline, recording, preview and export operations used by the app. Read an actual Codex tool trace ↗

Recording and editing run on your Mac. Connected AI clients may send requested transcripts and preview frames to their provider.

A FEW GOOD QUESTIONS

Before your
first take.

Is it free and open source?

Yes. The source is available under the MIT license. There are no artificial project or recording length limits. Available disk space and your Mac determine capacity. Local tools need no API key. Optional cloud providers and connected AI clients use their own billing or plans.

Which Macs can run it?

The current download targets Apple silicon Macs running macOS 15 or later. Node.js is included with the app. The public beta is Developer ID signed and Apple notarized.

Does my recording leave my Mac?

Recording, editing and Whisper transcription run locally. Model downloads need an internet connection. Optional cloud assistance sends the necessary transcript, edit context and your prompt to the provider you choose. External MCP clients may also send requested transcripts and preview frames to their own provider.

Can I edit after exporting?

Absolutely. Export produces a separate video. The project retains its original media, camera intervals, effects and captions for future edits.

What permissions does it need?

Screen recording, camera and microphone permissions depend on what you capture. Pointer movement works without Input Monitoring. Enable it for typing activity and precise click detection; typed text and key codes are never stored. macOS manages these permissions in Privacy & Security.

How do I get help or contribute?

Open an issue on GitHub, suggest an improvement, or emailinfo@serkanuslu.com. Contributions and thoughtful feedback are welcome.

YOUR NEXT GREAT VIDEO STARTS HERE

Go on.
Make your next take.

A small studio for the things you know.

Download for Apple silicon ↓

v0.1.2 · macOS 15+ · Public beta

Developer ID signed and Apple notarized public beta for Apple silicon.
Prefer to build it yourself?Get the source.