Video tools

2× and 4× WebGL Lanczos upscaling applied frame-by-frame to video clips. Queue multiple files, retry failures. No upload required.

Real-time webcam background removal. Virtual background modes: transparent, blur, solid, or custom image.

Apply 22 cinematic filter presets to any video. Adjust intensity. Export WebM. 100% private browser processing.

Remove jitter and shakiness from handheld footage with motion analysis and trajectory smoothing.

Drop a product video onto a lifestyle scene. Frame-by-frame AI background removal composites it onto 14 scene templates, exported as MP4 or WebM.

Turn one property photo into a cinematic 3D parallax fly-through. AI depth estimation drives a real camera move, exported as MP4 or WebM.

Strip creation date, GPS, and encoder info from MP4, WebM, and MOV video files.

Convert video between MP4 (AVC), WebM VP9, and WebM VP8 entirely in-browser.

Reduce video file size with High, Medium, Low, and Very Low quality presets.

Scale video to 4K, 1080p, 720p, 480p, or a custom resolution.

Add letterbox or pillarbox padding to change aspect ratio — 1:1, 9:16, 16:9, and more.

Rotate video 90°, 180°, or 270°. Fix sideways or upside-down footage.

Compare WebM VP9, WebM VP8, and MP4 file sizes side by side from one upload.

Transcribe speech to word-timed subtitles (SRT / VTT / burned-in). Whisper-base, multi-language detect.

Cut silent sections, filler words, or non-speech audio. VAD-driven, preserves natural pacing.

Remove background hum, fan noise, and AC drone from voice recordings. Real-time on most devices.

Split a mixed audio track into vocals and instruments. Demucs Small, ~5–15 s per 30 s clip.

Detect scene cuts via histogram + colour-cluster heuristics. Generates chapter markers and per-scene thumbnails.

Distil a long clip into a 15 / 30 / 60-second highlight cut. Combines scene detection + per-frame composition scoring + speech-segment heuristics.

Describe what happens across a clip — per-keyframe captions aggregated into a narrative, bullets, SEO blurb, social caption, or hashtags. Add an optional topic to ground the result.

Convert horizontal video to vertical (or square) by tracking the subject across frames and computing a smooth crop path. The "horizontal-to-TikTok" tool.

Apply a style transfer per-frame with seed-locking for temporal coherence. Experimental — 15 s/frame at 512 px.

Translate Whisper-generated subtitles into another language via the loaded WebLLM. Source language auto-detected.

Compress a long transcript into a 3-sentence summary, bullet points, or SEO description via WebLLM.

YouTube-style chapter markers: scene detection + per-scene caption + LLM titling.

Find the sharpest, best-composed frame across a clip. Sharpness + face-presence + composition scoring.

Best-frame + enhance (face restore + auto-grade) + optional text overlay. Output is YouTube / TikTok / Instagram ready.

Pick the highlight segment, encode as optimised GIF with loop polish. Auto-detects ideal segment when no in/out given.

Find seamless loop in/out points by frame-similarity matching. The "make this GIF actually loop" tool.

FFT-based onset detection — returns beat timestamps in seconds for music-driven editing.

Snap cut points to beats. Pair with scene detection for cuts that hit drops automatically.

Per-frame face detection + temporal smoothing so privacy-blurred faces don't pop in / out as detection wavers.

Same temporal smoothing as face-blur but for licence plates. Driving / dashcam footage prep.

Split a transcript by speaker. Useful for interviews and podcasts — labels every line "Speaker 1 / Speaker 2 / ...".

2x / 4x frame interpolation via classical bilinear blending — smooth slow-mo from any clip. 0 MB download.

Drive a character image with your webcam — the puppet follows your live motion. Classical motion-tracking, 0 MB download.

Animate a portrait's mouth in time with an audio track's loudness envelope. Classical lip-motion, 0 MB download.

Browser-native text-to-speech via Web Speech API. Pick from system-installed voices. 0 MB download.