Crop Video Online Free: Crop, Resize and Edit Videos

Upload a video file

Video files are uploaded by dragging and dropping local files or selecting media from cloud storage repositories. Standard browser editors accept inputs up to 100 MB on basic tiers, while scalable web architectures manage HTTP PUT payloads up to 2 GB (GitHub Docs, 2026; Google Studio Help, 2026). Large-payload handling is why processing speed, not just capacity, defines usable throughput. A 900 MB ceiling means very little if the queue takes eleven minutes.

To upload your video, drag the file into the web canvas or click the upload button and browse local folders. Common input formats include MP4, MOV, and WebM containers (Google Vids, 2026). Choosing a web-optimized container shortens the initial buffer delay while the browser renders the first frame.

Importing from cloud storage and direct URLs. Modern browser editors ingest media directly from cloud providers (Google Drive, Dropbox, Microsoft OneDrive) and from public URL endpoints. To import from cloud storage, select the provider icon in the upload modal, authorize account access through the OAuth prompt, and stream the file into the browser workspace without consuming local disk space. To import by link, paste a publicly reachable direct URL; the service fetches the remote file server-side, which avoids a double transfer on metered mobile connections. Cloud round-tripping also works in reverse: after cropping, the export can be written back to the same connected drive instead of the local download folder.

One practical caveat. Private or permission-gated links fail silently on many services, so set sharing to "anyone with the link" before pasting the address. If your source lives on a video host rather than a drive, download the master file first, because most croppers reject streaming manifests such as HLS .m3u8 playlists.

Select the crop frame or aspect ratio

Frame selection works two ways: drag the interactive bounding handles over the canvas, or apply a preset aspect ratio. A locked preset fixes horizontal and vertical proportions; a freeform preset opens up custom spatial boundaries (IMG.LY Docs, 2026; Adobe Help Center, 2025).

Position the crop frame video online selection boundary directly over the focal subject. Sounds obvious. It is also the step people rush.

Standard social presets lock proportions to 1:1, 9:16, or 16:9, and any competent browser video editor exposes an unlock toggle for custom geometry. Unlocking the ratio lets you drag side handles independently and crop frame of video online assets for bespoke web layouts. Mobile-first editors behave the same way; if your team already works inside a capcut video editor habit, the browser equivalent maps onto the same mental model of handles plus presets.

Manual pixel coordinate entry. For absolute precision, skip the drag handles and type numeric pixel coordinates (X, Y) alongside target width and height (W, H). Manual entry guarantees that cropped frames conform to strict UI placement specifications, ad-slot dimensions, and header banner requirements. A reliable working method:

Pause playback on the frame that contains the tightest composition. Cropping decisions made on a static reference frame beat decisions made mid-motion.

Enter W and H first (for example 1080 × 1920), so the frame size is fixed before positioning.

Enter the X and Y offsets to place the top-left origin of the crop rectangle.

Keep W and H even numberschroma-subsampled 4:2:0 encoding works on 2×2 pixel blocks, and odd dimensions force encoders to pad or round the frame.

Scrub the timeline once more to confirm the subject never leaves the fixed rectangle. A crop rectangle is static. The action is not.

Preview, edit and download the cropped video

Real-time previewing lets you verify focal alignment and spatial bounds before generating the final export. Lossless rendering passes unchanged frame segments straight through to export, avoiding needless re-encoding and quality degradation when source parameters match (NCH Software, 2026).

Review playback in the online video cropper canvas and check that key visual elements stay visible for the whole clip. Once the spatial adjustments look right, run the export command to crop your video and save the processed video file locally, or push it back to the connected cloud drive. Teams that need repeatable output rather than one-off edits should look at the api route instead, where the same offsets apply to a whole batch.

Crop the frame without changing video length

Resize and trim a video clip

Resizing scales spatial resolution down or up to fit destination dimensions; trimming sets start and end points along the timeline. Per W3C standards, spatial transformations scale and crop visual coordinates, while temporal trimming operates independently through mark-in and mark-out timestamps (W3C Media Capture and Streams, 2025).

When you need to crop and edit video online, pairing spatial scaling with temporal cuts is what actually shrinks total file size. Trim unnecessary intro or tail frames first and you reduce processing overhead before the spatial pass runs. Teams benchmarking editing stacks against one another can review the best free video editing software for feature-by-feature comparison of crop, resize, and trim behaviour, browse the wider AI Media Comparison Matrices for adjacent tool categories, and study end-to-end YouTube video editor workflows when publishing follows immediately after cropping. A crop cut video editor that handles both operations in one session also spares you a second export generation, which is where quality quietly leaks.

Make square videos for social media

Square videos (1:1, typically 1080×1080 pixels) maximise feed footprint on desktop and mobile alike. Published production guidelines recommend the square frame precisely because it accommodates footage originally shot in 16:9 or 9:16 and avoids letterboxing, provided you fill the frame rather than pad it (University of Glasgow Video Guidelines, 2020). Centering the subject during reframing keeps key actions visible without clipping during cross-platform syndication; where action sits near the edges of a 16:9 source, reposition the shot along the x-axis instead of simply scaling it (Frame.io, 2017).

Converting landscape sources to square requires that horizontal repositioning. Preset aspect ratio masks let editors crop a video online quickly while keeping symmetrical visual balance, and the same masks handle the reverse case when you crop my video online from a portrait original. Creators building assets from scratch instead of reframing existing footage can start from a free AI video generator that outputs native square and vertical dimensions, which removes the crop step entirely. Illustrated or animated formats behave the same way: a cartoon maker that renders to spec saves a reframing pass later.

Crop vertical video for TikTok and Instagram

Vertical video (9:16, 1080×1920 pixels) is the native standard for TikTok, Instagram Reels, and YouTube Shorts. Reframing horizontal footage means scaling content to fill a vertical canvas while keeping focal points inside central safe zones, clear of platform UI overlays (TikTok Safe Zone Guides, 2026).

When you crop and resize video online for vertical delivery, keep text overlays and facial features away from the bottom caption box and the right-side engagement icons. As a practical margin, treat the top ~10% and bottom ~20% of the 1080×1920 canvas as reserved interface space, and leave the right-hand column clear for the action rail. Teams producing vertical-first campaigns at volume often pair cropping with an AI video generator that renders natively in 9:16, which avoids lossy reframing altogether. For character-led vertical formats, a cartoon character maker with a 9:16 output preset does the same job upstream.

Crop video online on mobile browsers (iOS Safari and Android Chrome)

Browser-based croppers run inside iOS Safari and Android Chrome with no app installation, which matters when the source footage should never leave the phone. Open the tool's URL, tap the upload control, and choose Photo Library on iOS or Files / Gallery on Android; the picker hands the file to the browser directly.

Three mobile-specific constraints are worth planning around:

• Codec support depends on hardware. WebCodecs exposes only browser-native decoders (H.264, HEVC, VP8, VP9, AV1), and Safari's AV1 support relies on hardware decoding, so an AV1 source that opens on a desktop may fail on an older iPhone.
• iPhone HEVC captures may need conversion. Video recorded in "High Efficiency" mode is HEVC in a .mov container. If a cropper rejects it, set the camera to "Most Compatible" (H.264) or run it through a converter video step to MP4 before uploading.
• Background tabs suspend processing. Mobile browsers throttle inactive tabs, so keep the cropping tab in the foreground until export completes, and do not lock the screen mid-render.

For long clips on cellular data, trim first and crop second. Cutting the timeline before the spatial pass reduces the number of frames the phone has to decode and re-encode, which on a mid-range Android is the difference between 40 seconds and four minutes. If a mobile export stalls repeatedly, the AI Media Support and Troubleshooting notes cover the usual culprits.

Convert square video to circle online

Converting a square video to a circle format means centering the subject inside a 1:1 frame and applying a circular clipping path. CSS clip-path: circle() or an HTML canvas arc() combined with clip() restricts visible pixels to a circular radius (W3C CSS Masking Level 1; HTML Canvas 2D Specification).

To convert square video to circle online, position the source so the focal subject sits at the exact geometric midpoint. Applying the circular boundary crops the outer corner pixels and leaves a clean circular asset for profile embeds. Two practical routes exist in browser tools. The first is a mask preset, a one-click circle applied to the selected clip. The second is an overlay or picture-in-picture route, where the circular clip is layered over a background video or solid colour so the round frame reads correctly on any surface. If the tool exports without an alpha channel, use the overlay route and match the background to the destination page colour, or export to WebM with alpha where supported.

Adjust the circle frame around the subject

Adjusting a circle frame means repositioning the central origin and dragging corner control handles until the subject is enclosed. Interactive preview controls allow real-time alignment, so facial features or core objects stay centered inside the circular boundary (Apple Preview, 2026; Open Images, 2026). Where the editor exposes numeric fields, entering the exact area size, as Epson's preview-scan interface does for document dimensions, is far more repeatable than dragging handles across a batch of clips.

When you use a circle crop video online tool, scale the circular boundary until the subject fills roughly 70% of the internal frame. That spatial margin prevents an uncomfortably tight crop around faces during mobile playback. If the subject moves, apply keyframes to the mask position so the circle tracks the action instead of clipping it.

Oval, rounded and custom SVG shape crops

Advanced spatial masking goes past circles into ovals, rounded rectangles, and custom geometric SVG paths. Custom clipping masks let video overlays match complex UI layouts and promotional templates without hard rectangular edges. Three implementations are available in browser workflows:

• Ellipse clipping. clip-path: ellipse(50% 35%) produces an oval frame from a rectangular source without pre-cropping to 1:1.
• Rounded rectangles. border-radius or clip-path: inset(0 round 32px) softens corners for card-style embeds and dashboard widgets.
• SVG masks. Referencing an <svg> <mask> from mask-image supports arbitrary shapes (hearts, stars, flowers, badges, brand marks) plus gradient-based soft edges that hard clipping cannot produce.

The distinction matters technically: clip-path performs hard-edge geometric clipping, while mask and mask-image use alpha or luminance data and therefore support feathered fades (W3C CSS Masking Level 1). Template-driven editors expose the same capability under labels like "frames" or "shape presets", which is functionally an SVG mask with decorative borders attached. A cartoon maker app with sticker frames uses exactly this mechanism under the hood.

Crop MP4 video online

Cropping MP4 video means decoding H.264 or H.265 (HEVC) streams, applying spatial cropping offsets, and encoding the target rectangle. H.264 bitstreams use explicit frame_crop parameters, while H.265 relies on conformance cropping windows to define visible frame boundaries (ITU-T H.264 / H.265 Standards, 2026).

Research shows that optimising bitrate-resolution pairs with XPSNR metrics achieves an average 5.84 dB PSNR improvement while reducing encoding time by 44.43%, and cuts decoding time by 65.46% at the same bitrate relative to a baseline VVenC configuration (IEEE ICIP, 2024). To crop mp4 video online, upload the file, adjust the spatial frame boundaries, and export with tuned bitrate settings to hold visual quality. The same path covers people searching for crop mp4 video free online or crop mp4 video online free: the operation is identical, only the tier limits differ.

Choose a format for download

Exporting a cropped video means selecting an output container to match the destination. MP4 with H.264 video and AAC audio offers universal web compatibility, MOV supports high-bitrate Apple workflows, and WebM delivers efficient web playback with VP8, VP9, or AV1 payloads (AWS MediaConvert Docs; Azure Media Services). Reviewing the best AI video generator options is a fast way to check which export containers a given stack actually supports before you commit a batch to it, and the AI Media Calculators help you size storage and bitrate for that batch in advance.

When you crop a mp4 video online, choosing MP4 as the final export format minimises cross-device playback errors. HEVC in MP4 halves file size on modern hardware but fails on older devices. MOV output in managed pipelines is typically limited to AVC (H.264), MPEG-2, or Apple ProRes, and MOV inputs must be self-contained rather than reference-linked, which trips up editors who hand off Final Cut projects instead of flattened masters.

What is included in free video cropping

Free video cropping tiers usually include basic frame adjustment, standard social presets, and 1080p export at no upfront cost. Some platforms offer unconditional watermark-free exports; others restrict clean exports to projects built exclusively with free stock assets (Microsoft Clipchamp, 2026; Canva Terms, 2026). Open-source desktop alternatives such as OpenShot advertise unlimited tracks with no watermarks, subscriptions, or project limits, and that is the reference point against which browser free tiers should be judged.

Anyone planning to crop a video for free online should verify export conditions before editing, not after. Tools that genuinely offer crop a video online no watermark capability keep branding overlays off finished assets, which is a material concern for paid distribution: a competitor's watermark on a promoted asset is a brand problem and a compliance problem at once. If your team is comparing paid tiers rather than free ones, the AI Media Pricing overview sets out where per-minute and committed-throughput models start to make sense.

Organizations assessing licensing exposure across their media stack should review the commercial use of AI image generators and the broader AI Media Commercial-Use documentation, because the same rights questions, ownership of output, permitted resale, attribution duties, apply to processed video assets. Where output provenance is disputed, the ongoing AI Litigation and case tracker is the better starting point than vendor marketing copy.

Data retention and privacy. Free browser-based processing environments enforce data-wiping schedules of varying rigour. Uploaded media and temporary render buffers are typically deleted from edge processing servers within 60 minutes of session termination, and several free croppers publish this explicitly ("files are automatically deleted one hour after upload"). Where retention is not published, assume indefinite storage and act accordingly.

Free tier comparison: watermarks and limits

Figures reflect published vendor terms and comparison reporting for 2026 and change frequently. Confirm the current cap and watermark rule at the point of upload.

Large video files and quality after cropping

Cropping large video files preserves original resolution, frame rate, and bitrate only when export parameters match source settings. Re-encoding large files means managing compression trade-offs, since unoptimised re-compression introduces visible spatial artifacts (Vimeo Compression Guidelines, 2026). Understanding how a video compressor balances bitrate against perceived quality is the fastest way to avoid the classic mistake: cropping a 4K master and exporting it at a 720p-class bitrate.

When you process footage with a crop large video online free utility, or a crop big video online tool aimed at multi-gigabyte masters, lock constant frame rate parameters to maintain audio-video synchronization. Published bitrate targets give a workable floor: roughly 5 to 10 Mbps for 720p, 10 to 20 Mbps for 1080p, and 30 to 60 Mbps for 4K. Upload architecture matters too. Chunked transfer is required above roughly 100 MB in several managed pipelines, and asynchronous handling kicks in above 20 GB (Cloudinary Docs, 2026).

File retention and encryption

Managed processing platforms document encryption and regional residency explicitly. Azure Content Understanding, for example, states that uploaded files are temporarily encrypted and stored in Azure Storage in the same region as the resource before onward processing. Consumer croppers rarely publish equivalent detail; the ones that do usually commit to automatic deletion within one hour of upload. Where neither encryption at rest nor a deletion window is stated, treat the upload as permanent and public in your risk model.

Distribution settings do not travel with the file, either. Vimeo notes that publish-to-social transfers only the video file: platform privacy settings, interaction tools, captions, and thumbnails are not carried over. A clip that was restricted on one host can therefore become fully public on another, without anyone changing a permission.

Enterprise readiness checklist

Before approving a browser cropper for corporate footage, confirm the following in writing:

1. Retention window.Is deletion automatic, and within what interval after session end?
2. Encryption.Is media encrypted in transit (TLS) and at rest, and in which region?
3. Model training.Are uploaded files excluded from AI model training and human review?
4. Compliance posture.Is there a SOC 2 Type II report, ISO 27001 certificate, GDPR data processing addendum, and a documented sub-processor list?
5. Access control.SSO/SAML, role-based permissions, and audit logs showing who exported what?
6. Rights warranty.Do the terms require you to warrant rights in uploaded content, and what licence do you grant the provider?
7. Watermark and export rights.Is commercial use of the output permitted on the tier you are actually using?
8. Deployment options.Is a private-cloud, VPC, or on-premise processing path available for restricted material?

Two of these tend to fail quietly. Model-training exclusion is often absent rather than denied, and sub-processor lists are frequently published once and never updated.

Free web tool vs enterprise API or private cloud

For high-volume reframing, programmatic cropping is the pragmatic answer. A single API call applies fixed X/Y/W/H offsets and an aspect-ratio preset across an entire library, with deterministic output and a log trail that no manual browser session can produce. One caveat on the evidence here: vendor-published SLAs and residency terms differ enough that this table should be read as a shape of the decision, not a substitute for reading the contract.

How cropping is signalled in H.264 and H.265

ITU-T H.264 specifies frame_crop_left_offset, frame_crop_right_offset, frame_crop_top_offset, and frame_crop_bottom_offset to define the decoded output rectangle, with crop units dependent on the chroma format and frame_mbs_only_flag. H.265/HEVC conformance text instead defines a conformance cropping window, so the presented output size can be signalled without altering coded samples. The two standards expose cropping through different syntax, which is exactly why a "crop" in one pipeline may be metadata-only and in another a full re-encode.

Lossless passes and when re-encoding is forced

Vendor documentation for lossless export is unambiguous about its conditions. The source must be H.264 or MPEG-4, input and output format must match, dimensions, profile, levels, and extradata must be identical, and containers are limited to .avi, .mkv, .mov, .mp4. Visual effects are not supported, affected segments are re-encoded, and if fewer than 10% of the sequence qualifies, the entire sequence is re-encoded (NCH Software, 2026). A spatial crop changes dimensions by definition, so cropping is always a re-encode of the affected region. The lossless path applies only to untouched temporal segments.

Avoiding blur and compression artifacts

For archival or scientific-grade recompression, published guidance recommends conservative settings, for example CRF 10 to 12 with a slower x264 preset, trading encode time for near-transparent quality (NASA PDS H.264/MP4 guidance). Feature-level differences between consumer editors, including which of them expose CRF at all, are catalogued in the capcut video editor features reference and the accompanying capcut video editor overview.