ReelMimic Style-Mimicking Video Studio

What does this agent do, and when should you use it?

ReelMimic is a locally run, browser-operated studio that turns a reference video into an original animation that borrows its style. You drop in a file, a phone recording or a YouTube link, describe what you want, and it first breaks the reference down into shot count, shot lengths, BPM, transitions, colours, framing and camera moves. It then presents a plan of storyboard, characters, assets and style frames for you to discuss and approve. Production is split across up to 6 AI agents working in parallel, and every finished shot is handed to a separate agent for review, with fixes backed by before-and-after screenshots. It runs entirely through your own Claude Code or Codex account, renders 2D animation through seven drawing engines, and can push the finished cut to social platforms via Upload-Post.

After ingesting a dropped file, phone recording or YouTube link, ReelMimic derives shot count, shot lengths, BPM, transitions, colours, framing and camera moves, then generates a plan covering storyboard, characters, assets and style frames, which you can revise through the side chat (text or screenshots). On "Approve and start", work is split into units executed by up to 6 concurrent agents; each completed shot goes to a fresh agent for review, and every fix must ship with before-and-after screenshots that the reviewer checks. The Production line tab shows where each character and each part of the video stands. Seven drawing engines are available — vector/motion graphics (built on HyperFrames), hand-painted watercolor (built on painted-animation), crayon picture book, pixel art, paper cut-out stop-motion, whiteboard doodle and anime cel — and each style is a Markdown file, so adding one needs no code. Assembly relies on FFmpeg, the Chrome renderer and a Python toolchain, and a finished project can be published with npm run post-video -- <project id> --platforms tiktok,instagram,youtube, which only actually sends with --send.

1. A solo creator who wants the pacing and camera moves of a phone clip applied to their own original characters
2. An indie music-video team needing quick hand-painted or crayon-style animation
3. A marketer replicating the cadence and transitions of an existing brand animation for a new product film
4. A developer who wants the whole pipeline to run locally on their own Claude Code or Codex quota without uploading source footage
5. A tinkerer adding custom drawing styles or rendering engines, describing styles in Markdown
6. A content operator who needs to push the finished animation to TikTok, Instagram and YouTube

How do you install or deploy this agent?

Requirements: Node.js 22.18+, Python 3.10+, FFmpeg, Chrome, and a logged-in Claude Code or Codex CLI.

git clone https://github.com/edenfunf/reelmimic.git && cd reelmimic
./install.sh      # on Windows, double-click install.bat
./start.sh        # on Windows, double-click start.bat

Then open <http://localhost:4318>. The install script checks your setup and can be re-run any time:

cd app && npm run doctor

API keys and tool paths live outside the repo in ~/.reelmimic/secrets.json; the format is shown in secrets.example.json:

{
  "YATING_KEY": "Yating's Taiwanese Mandarin voices, for narration",
  "PIXABAY_KEY": "",
  "FREESOUND_KEY": "",
  "FFMPEG_DIR": "",
  "CHROME_PATH": "",
  "CODEX_BIN": "",
  "PYTHON": "",
  "CODEX_SANDBOX": "danger-full-access",
  "BUILDERS": 6,
  "MAX_AGENTS": 12,
  "PORT": 4318
}

How do you use this agent?

On the home page, drop in a reference video or paste a link, describe what you want, and pick Claude Code or Codex. Wait for the breakdown and plan, then request changes in the chat panel on the right; if the plan asks for something (lyrics, say), supply it or skip it, and hit "Approve and start". During production, the Production line tab tracks each character and each part of the video along with review screenshots. When it finishes, scrub to any second and leave a note, then send all notes at once.

To publish a finished cut that has passed final review:

npm run post-video -- <project id> --platforms tiktok,instagram,youtube
npm run post-video -- <project id> --platforms tiktok --at 2026-10-05T18:00 --tz Europe/Madrid --send

Without --send it is a dry run and nothing leaves your machine; re-running the same video will not post it twice.

What are this agent's strengths and limitations?

How does this agent compare with similar options?

Key facts side by side with the most closely related agents.

How does FollowAgents rate this agent?

FAQ