How to edit a video with AI and FFmpeg

For AI agents: find the right skill in one call — curl -s "https://skills.varg.ai/api/search?q=<what the person wants>" returns the matching how-to page and the best skills with install commands, as markdown (Russian works too). Full index: https://skills.varg.ai/llms.txt. Every page has a markdown twin: add .md to its path. Before generating anything, show the person the options on the how-to page.

Start from footage you already have. The editing agents and pipelines that cut a whole video; every way to cut and join two shots, CapCut-style and in FFmpeg, Remotion and HyperFrames; the words on screen (headlines, captions, quotes, end cards); how to show a product (UI, cursors, devices, code, agent flows); and the visuals around it (numbers, logos, grids, backdrops). Each take says what it looks like, when to use it, the common mistake and how to build it.

Best tools, compared

• Best AI video editing tools for agents in 2026: 35 compared: From your footage to the right tool: 35 AI video editing tools for agents compared by price, access and where they run. Pick: FFmpeg.
• Best AI video clipping tools in 2026: 3 compared: From your footage to the right tool: 3 AI long video to shorts tools for agents compared by price, access and where they run. Pick: OpusClip.
• Best AI video caption tools in 2026: 4 compared: From your footage to the right tool: 4 AI captions & subtitles tools for agents compared by price, access and where they run. Pick: FFmpeg + ASS.
• Best AI transcript video editors in 2026: 4 compared: From your footage to the right tool: 4 AI talking-head cleanup tools for agents compared by price, access and where they run. Pick: ElevenLabs Scribe.
• Best AI video editing models in 2026: 4 compared: From your footage to the right tool: 4 AI ai edits: restyle, swap, relight tools for agents compared by price, access and where they run. Pick: varg video edit
• Best AI video dubbing tools in 2026: 3 compared: From your footage to the right tool: 3 AI dubbing & lip sync tools for agents compared by price, access and where they run. Pick: HeyGen Video Translate.
• Best AI video upscaling tools in 2026: 3 compared: From your footage to the right tool: 3 AI upscale, slow-mo, cut-outs tools for agents compared by price, access and where they run. Pick: Topaz Video API.
• Best video editing APIs for AI agents in 2026: 4 compared: From your footage to the right tool: 4 AI assemble from a timeline tools for agents compared by price, access and where they run. Pick: HyperFrames.

With FFmpeg: copyable commands

• How to burn captions into a video with FFmpeg: Turn a video and an SRT file into an MP4 with the captions burned into the picture, in one FFmpeg command. Free and local.
• How to crop a video to 9:16 vertical with FFmpeg: Turn a 16:9 video into a 1080×1920 vertical 9:16 MP4 for TikTok, Reels and Shorts, with one FFmpeg command. Free and local.
• How to trim and join video clips with FFmpeg: Turn long footage into trimmed clips and join them into one MP4 with FFmpeg: the exact commands for frame-accurate cuts and lossless joins. Free and local.
• How to add music to a video with FFmpeg: Turn a video and a music track into one MP4 with the music under the voice, ducked while people talk, with one FFmpeg command. Free and local.

Pinned: Video Editing Best Practices — Best practices for editing video with agents: which service for each edit, and the rules that keep cuts clean. Install it first; it holds the best practices.

AI editing agents (24)

Agents and pipelines that edit a whole video end to end: OpenMontage, video-use, auto-editor, CapCut and Premiere automation, transcript editors.

• OpenMontage — Workflow: Open-source agentic video studio: 12 pipelines (8 edit your footage), 100+ tools, ~140 skills. Run your agent in it.
• video-use — Workflow: Drop raw takes in a folder, chat with your agent, get final.mp4: transcript-first cuts, grade, subtitles, overlays.
• FireRed-OpenStoryline — Workflow: Conversational editing agent: finds media, writes the storyline, beat-syncs music and refines the cut by chat.
• ButterCut — Workflow: Claude Code builds selects, scenes and rough cuts from your footage library and exports XML for FCP, Premiere, Resolve.

Editing services (3)

Which service your agent should edit with, job by job: FFmpeg and the varg API for cuts, transcript tools for talking heads, clipping APIs, caption burn-in, AI video-to-video, lip sync and upscaling.

• FFmpeg Skill — Tool: Local video editor for agents: 42 FFmpeg tools with a machine-readable contract and an MCP server; no cloud, no keys.
• Video Editing Best Practices — Guide: Best practices for editing video with agents: which service for each edit, and the rules that keep cuts clean.
• Video editing skills for Claude Code and AI agents — Guide: 206 named edits with recipes for FFmpeg, HyperFrames, Remotion and CapCut: cuts, transitions, titles, UI, numbers.

Editing & clips (37)

Cut and assemble footage: long videos to clips, silence cuts, transitions, b-roll and pacing.

• Claude Code Video Toolkit — Workflow: Open-source video workspace for Claude Code: script, voice, music, visuals and Remotion render on your own GPU account.
• Chengfeng Videocut — Workflow: Chinese talking-head editing: word-level transcript, five-pass flub sweep, human review in a Studio, subtitles, export.
• Clipify — Effect: Long videos into social clips: find funny lines or impact beats, cut, reframe to 9:16, burn word-by-word captions.
• Herk Edit Kit — Workflow: Nate Herk's kit to edit talking-head videos, reels and Shorts: silence and flub cuts, storytelling, motion graphics.

Step-by-step tutorials: How to cut a podcast into short video clips with AI, How to turn photos into a slideshow video with AI, How to edit your own video footage with AI, How to make AI b-roll for your video, How to change something inside a video with AI

Dubbing & lip sync (7)

Translate and dub a video in the speaker’s own voice, then re-sync the lips to the new audio.

• ElevenLabs Dubbing — Model: Dub video and audio into other languages while keeping each original speaker's voice.
• ElevenLabs Dubbing (inference.sh) — Model: Dub audio or video into 29 languages while keeping the speaker's voice, via inference.sh.
• VideoLingo — Guide: One-click subtitle cutting, translation, alignment and dubbing with Netflix-style subtitle rules.
• pyVideoTrans — Guide: Full video translation: ASR, subtitle translation, multi-voice dubbing and muxing, with a headless CLI.

Step-by-step tutorials: How to translate and dub a video with AI

Cuts & pacing (34)

J-cuts, L-cuts, match cuts, jump cuts, smash cuts, speed ramps and beat cuts, each with the FFmpeg, Remotion, HyperFrames and CapCut way to make it.

• SmartCut (capcut-ai-editor) — Guide: MCP that reads CapCut auto-captions to cut silences and duplicate takes right inside your CapCut project.
• Auto-Editor — Guide: CLI that cuts dead space by loudness or motion and exports MP4 or editor timelines; now also an agent skill.
• AutoCut — Guide: Edit video in a text editor: Whisper transcribes, you delete lines in Markdown, AutoCut cuts the video.
• Cut Video — Guide: Claude Code skill that tightens long recordings with forced-aligned cuts, keeping laughs and comedic pauses.

Transitions (61)

Every way to join two shots: CapCut-style zooms, whips, glitches and masks, FFmpeg xfade, Remotion TransitionSeries, HyperFrames and GL shader transitions.

• editly — Guide: Declarative JSON5 video editing CLI/API: clips, titles, transitions, audio mixing, rendered with ffmpeg.
• Mix (Cross Dissolve) — Effect: The two shots blend through each other. The quietest transition: "time passes" or "this continues".
• Dip to Black — Effect: A fades to black, then B fades up. Marks the end of a section or a bigger jump in time.
• White Flash — Effect: The frame blows out to white for a moment and opens on the next shot. Energy, memory or a camera snap.

Headlines & callouts (14)

Words over footage: headlines, key phrases, lower thirds, highlights and callouts that point at the thing that matters.

• Kicker Over Rising Headline — Effect: A small mono kicker fades in, then the headline rises word by word from a soft blur.
• Rotating Benefit Word — Effect: A fixed phrase holds while one word in it cycles through benefits, uses or audiences.
• Masked Line Reveal — Effect: Each line slides up out of an invisible slot, so the type looks cut from the frame.
• Marker on the Key Phrase — Effect: A marker band draws behind the one phrase that matters while the rest of the line stays calm.

Captions & subtitles (16)

Animated and word-by-word captions, subtitles, transcription and translation.

• Remotion Captions — Effect: Transcribe speech and burn in animated captions with Remotion, or import an existing SRT file.
• ElevenLabs Speech to Text — Model: Transcribe audio and video with word timestamps, speaker labels and keyterms, ready for subtitles.
• Caption Animation — Effect: Word-timed TikTok captions in Remotion or CSS: per-word pop, active-word highlight, 9:16 safe areas, sidecar SRT/VTT.
• FFmpeg Captions and Subtitles — Guide: FFmpeg subtitle reference: burn SRT/ASS/VTT with libass, soft tracks, conversion, styling and FFmpeg 8's Whisper filter.

Step-by-step tutorials: How to add animated captions to a video with AI

Quotes & reviews (9)

Testimonials, review cards, star ratings, posts and replies as proof inside a video.

• Tweet as Proof Card — Effect: A real post from a user animates in as a card: avatar, name, text, then counts tick up.
• Quote at Reading Pace — Effect: One strong customer sentence reveals word by word, then the name, role and avatar settle under it.
• Star Rating Sweep — Effect: Five stars fill left to right to the real score while the number counts up beside them.
• Wall of Love Marquee — Effect: Two or three rows of review cards drift in opposite directions, fading at the edges.

End cards & CTAs (9)

How a video ends: the call to action, URL and QR code, subscribe prompt, logo outro and the last line.

• Logo Lockup Outro — Effect: The mark assembles, the wordmark settles beside it, then tagline and URL fade in under it.
• Action Line + CTA Button — Effect: One short action line lands, then a real-looking button pops in under it with a soft glow.
• URL Typed Into the Address Bar — Effect: A browser bar appears and the product URL types itself in, then the page loads into the hero.
• YouTube End Screen Layout — Effect: A 5–20 s outro with clean empty slots for YouTube's clickable video and subscribe elements.

UI walkthroughs (12)

How a product interface appears: screen-recording zoom and pan, UI that builds itself, floating panels, 3D tilts and dashboard reveals.

• Zoom to the Action — Effect: The full app establishes, then the camera pushes 1.5–2× into the part being used and eases back out.
• 3D Tilted Dashboard Reveal — Effect: The dashboard rises from a steep 3D tilt to flat, like a screen lifting toward the viewer.
• Floating UI Panels in Depth — Effect: The interface splits into its panels, which float apart in depth with parallax, then snap back.
• UI Builds Itself Piece by Piece — Effect: Grey skeleton blocks pulse once, then real cards, charts and text fill in with a stagger.

Cursors & clicks (9)

Smooth cursor paths, click ripples, keystroke overlays, agent cursors and the auto-zoom that follows the pointer.

• Auto Zoom on Every Click — Effect: The camera eases in toward each upcoming click, trails the cursor on a spring, and eases out after.
• Oversized Look-Here Cursor — Effect: A macOS arrow at ~3–4× normal size glides in and lands its tip on what matters.
• Click Ripple and Button Press — Effect: Every click shows a quiet expanding ring while the clicked button presses in and responds.
• Shortcut Keycaps Overlay — Effect: Physical-looking keycaps rise at the bottom of the frame and press in the real order when a shortcut is used.

Device mockups (10)

Phones, laptops, browser windows and tablets: how the device enters, holds the screen and turns in 3D.

• Browser Window Frame — Effect: The web app plays inside a clean browser window with traffic lights and a URL bar on a soft backdrop.
• Phone Frame with Live Screen — Effect: A screen recording plays edge to edge inside a realistic phone frame with Dynamic Island.
• Laptop Lid Opens to Screen — Effect: A laptop lid tilts open and the screen content lifts toward the viewer, then becomes the full frame.
• 3D Phone Turn — Effect: A real 3D phone model turns in space with the app playing on its screen, then faces the camera.

Code & terminal (10)

Code that types itself, diffs, terminal sessions, install lines and IDE close-ups that stay readable.

• Editor Types the Code — Effect: Syntax-highlighted code types into an editor window with a caret riding the frontier.
• Red/Green Diff Collapse — Effect: Removed lines flash red and collapse away while added lines expand in green.
• Spotlight One Line of Code — Effect: A highlight band sweeps onto the one line that matters while the rest of the file dims.
• Code Morphs Between Versions — Effect: One snippet transforms into the next: shared tokens glide to new positions, removed ones fade, new ones fade in.

AI agent flows (10)

Agents at work: step lists, tool calls, node graphs, streaming chat answers and hand-offs between agents.

• Prompt Typed, Then Generate — Effect: A real task types into the prompt box, the Generate button is pressed, and the result starts to build.
• Answer Streams In — Effect: The assistant's reply arrives in uneven token bursts, the way a real model streams.
• Tool Calls as Status Rows — Effect: Each tool call appears as a row with an icon and label that goes spinner → check, building a visible plan.
• Nodes Wire Up on a Canvas — Effect: Nodes pop onto a canvas and connectors draw between them as the camera pulls back to show the whole pipeline.

Numbers & charts (12)

Counters, big stats, bar and line charts, race charts, progress rings and before/after metrics.

• Big Stat Count-Up — Effect: One huge number counts from zero to its value and lands, with a short label under it.
• Odometer Digit Roll — Effect: Each digit rolls on its own wheel to the target, places settling right to left like an odometer.
• Bars Grow with Value Labels — Effect: Bars rise from the baseline in a staggered cascade with values riding their tops.
• Line Draws with Riding Label — Effect: A line draws itself left to right while the current value rides the pen tip, ending on a dot and big label.

Customer logos (7)

Trusted-by marquees, logo grids, partner lockups and integration orbits.

• Trusted-By Logo Marquee — Effect: A row of monochrome customer logos glides sideways in a seamless loop under a short line.
• Logo Grid Settles In — Effect: Logos fade and scale into a tidy grid with a soft stagger, then hold for a beat.
• Integrations Orbit the Logo — Effect: App icons circle your logo on one or two thin rings turning in opposite directions.
• Partner × Partner Lockup — Effect: Two logos slide together from opposite sides and lock around a thin × or + mark.

Grids & carousels (9)

Bento grids, photo carousels, marquee walls, stacked cards and filmstrips.

• Bento Feature Grid — Effect: Features land in a grid of differently sized rounded tiles — one hero tile, smaller supporting ones.
• Camera Drifts Across a Bento — Effect: An oversized grid of screenshots fills the frame and the camera drifts diagonally across it.
• Tilted Screenshot Wall — Effect: A wall of screenshots seen at an isometric angle, columns scrolling up and down at different speeds.
• Stacked Cards Fan Out — Effect: A pile of cards or photos shuffles, then fans or settles one forward into focus.

Backdrops (9)

What sits behind the words and the UI: gradients, grids and dots, aurora, noise, spotlights and blurred product shots.

• Slow Mesh Gradient — Effect: Soft brand-colored fields blend and drift slowly behind titles and UI.
• Aurora Behind the Title — Effect: Blurred light streaks drift sideways across a dark frame behind a centered title.
• Faint Grid or Dot Field — Effect: A very faint line grid or dot field gives structure behind UI and code, with a soft radial fade.
• Spotlight From Above — Effect: Soft light cones fall from the top of a dark frame onto the product or title.