Media tools
Sume media tools by API: trim, filter, captions, frames, inspect, speech-to-text, upscaling, background removal, transitions, and Timeline assembly.
Start with Video editing API limits: Sume media tool caps in one table
Shopify video limits: 120 fps and WebM allowed, so Sume MP4s fit
Shopify takes video up to 1 GB, 10 min, 120 fps and 4096 px per side. Sume's 24, 25, 30 or 60 fps MP4 at up to 2160 px sits inside all of those limits.
Shopify Magic edits at 1 megapixel: when to generate larger
Shopify media generation defaults to about 1 megapixel, resizing inputs to fit. Sume's Image API accepts custom sizes up to 3840 px and 8,294,400 pixels.
Product turntable ad: which Sume video models take two frames
For a product turn, give a video model the front and back photo as first and last frames. Six Sume rows take both; Grok Imagine takes a first frame only.
Dry-run a Sume video crop with video-filter check before a batch
POST /v1/video-filter/check validates a crop program without a job, credits or an encode. Run it first, then crop a batch of ad clips for new ratios.
YouTube Shorts custom thumbnail: Studio on a computer, 9:16 still
A custom Shorts thumbnail is 9:16, minimum 640 px on the short side, and can only be added in YouTube Studio on a computer. Export one still with Sume.
YouTube thumbnail 2 MB on mobile vs 50 MB on desktop: video still
YouTube custom thumbnails are limited to 2 MB on mobile and 50 MB on desktop, JPG or PNG. Cut a still from a Sume clip and choose JPEG to stay under 2 MB.
Lip-sync a saved avatar with H3 Max: avatar_handle, not image_url
Use a ready Sume avatar as the face for MiniMax H3 Max Lip Sync: send avatar_handle plus Sume-hosted audio, 5-14.8 seconds, and see the price at 768p.
Six product photos as references: which Sume video models accept them
Six reference photos fit Gemini Omni Flash, Wan 3.0, MiniMax H3 and H3 Max on Sume. Kling 3 and Grok Imagine Video take no references.
Check an ad clip's aspect ratio with Sume video inspect stills
Sume video inspect returns stills with width and height. Send max_edge at the source edge, read the ratio, and catch a wrong-ratio clip before upload.
YouTube podcast thumbnail: 1:1, 10 MB on mobile, a square Sume still
YouTube podcast playlist thumbnails are 1:1, with a 10 MB limit on mobile and 50 MB on desktop. Make a square cover with the Sume images API at 1:1.
Does Kling motion control make a photo talk? Where sound comes from
Kling 3.0 motion control copies movement from a reference video onto a still and takes no audio input. For speech from audio, use Fabric or H3 Max on Sume.
Timeline output fps for a Short: 24, 25, 30 or 60, and the warning
Timeline 1.0 output.fps accepts 24, 25, 30 or 60. Leave it out and the render follows the sources. Read output_fps_resamples_sources before you upload.
Trim an AI clip's slow intro for a Short: keyframe cut or exact cut?
Sume video-trim has two precisions: exact re-encodes to the frame, keyframe is a stream copy that can start a GOP early. Which one to use before a Short.
Fade a Short in and out with Timeline 1.0: fade seconds and limits
Timeline output.fade_in_seconds and fade_out_seconds take 0 to 5 s, and their sum must fit the output. The edge_fades_exceed_output refusal, with a request.
Dim a bright 9:16 clip before captions: video-filter dim at 0.45
Sume video-filter's dim op multiplies luma by an amount above 0 up to 1. A free check validates it, then one $0.02 job makes the clip easier to caption.
A 25-minute episode into eight teaser clips: four cents of audio
Detach a 25-minute video in two ranges under the 900-second cap, then split each into four teaser ranges: four Sume jobs, 4 cents, eight audio files.
Which Sume media step caps your clip: 1800, 900 or 300 s
Trim takes a 1800 s source, video filter and video frames stop at 300 s, compose at 300 s, Timeline at 1800 s. One table of every length cap.
Keep a whole 16:9 shot in 9:16: blur fill instead of cropping
Cropping loses 68% of a 16:9 frame. Timeline fit blur keeps the full shot centered over a blurred copy. How it works on Sume and when to prefer it.
Reframe 16:9 to 9:16: crop fractions for left, center, right
A 16:9 to 9:16 crop is a width of 0.3164 of the frame. Use x 0, 0.3418 or 0.6836 for a left, centered or right subject in Sume video filter.
Hand an AI clip to an editor: MP4, PNG frames and WAV on Sume
What Sume can deliver for a finishing editor: MP4 clips, lossless PNG stills at source size, and sample-exact WAV audio. No EXR or HDR is documented.
Color-grade AI video on Sume: lut3d is off, tone filters stay
Sume video filter has a dim op, a crop op and an allowlisted ffmpeg filtergraph. lut3d is not on it. What a grade can and cannot do, per the docs.
Pick the cleanest last frame: sample 24 stills before chaining
Before chaining AI clips, sample up to 24 stills with Sume video frames (fps up to 2) and choose the best one as the next first_frame, not just the last.
Camera orbit and zoom transition prompts for Omni first and last frame
Google says Omni 1.1 Flash handles orbits, zoom transitions and loops between two keyframes. Prompt patterns and costs for the Sume API, from $0.19 per draft.
Alexa skill audio from Sume music: 48 kbps MPEG-2 re-encode step
Alexa SSML audio wants MPEG-2 mp3 at 48 kbps and 16000, 22050 or 24000 Hz. Sume documents no music bitrate, so probe the file and re-encode it yourself.
Assert an edited image has the source's pixel size: Python check
A short Python check that downloads the source and the edited file and fails if the pixel size differs. Needed on Sume whenever you send an aspect_ratio.
audio_source_in_requires_single_spine: source_in needs one voice file
audio.source_in only works with one audio.url. With audio.parts, set source_in on each part instead; the Sume hint is use_part_source_in.
Cut a 90 s music window: timeline-audio split and range end errors
Use timeline-audio split with ranges [{start, end}] to cut a window. end must exceed start or you get audio_range_end_before_start; omit end for the rest.
Discord soundboard clip from Sume music: 5.2 s, 512 kb, mp3 or ogg
Discord's soundboard accepts mp3 or ogg up to 512 kb and 5.2 s. Cut a Sume track with timeline audio split to mp3 and check the size; wav will not fit.
Godot MP3 loop: begin point only, so loop a Sume track as wav
Godot gives MP3 and Ogg only a loop begin point, no loop end. For a Sume BGM loop, cut a sample-exact wav with timeline audio split and set the loop on import.
Link card 1200x628 from an Ideogram 4.5 edit: ask 2:1, crop 63 px
Ideogram 4.5 on Sume has no 1.91:1 ratio. Ask for 2:1 (1408x704 on a 1K edit), then crop about 63 px of width with Pillow to reach 1200x628.
invalid_segment_timing: the two causes in a 3-minute Short timeline
invalid_segment_timing means a slot start is not after the previous start, or a slot ends over 0.5 s past audio.duration_seconds. Check starts, then ends.
Is a reference Short silent? Reference ingest audio gate at -60 LUFS
Reference ingest marks audio.silent when integrated loudness is -60 LUFS or lower. Silent means plan new music and skip STT; the manifest also reports beats.
reference-ingest source_no_video_stream: audio-only file as input
source_no_video_stream means the reference file has no video track, such as an m4a. Send a clip with picture, or use audio detach to work with the sound.
reference_ingest_stt_required: speech.language_code without STT
speech.language_code and duration_seconds only apply when the read transcribes. Set speech.allow_billed_stt true, use reference_remix, or drop them.
Reference ingest text_tracks is not caption coverage: 5 sampled frames
text_tracks come from OCR on deduplicated frame states, at most five frames, not every caption. To know if a Short is captioned, check frames yourself.
silent_audio_takes_no_* in Sume timeline: BGM-only slideshow fix
audio.mode silence refuses url, parts, gain_db and source_in. Put gain and fades on the soundtrack bed, which is allowed over silence.
soundtrack_fade_exceeds_output: a bed fade longer than a short clip
Sume rejects soundtrack.fade_out_seconds longer than audio.duration_seconds. The fade also caps at 10 s. For a 6 s Short, use a 1 to 2 s fade.
Sume audio_url_and_parts_exclusive: pick one spine for a VO Short
Sume timeline rejects audio.url together with audio.parts. Send a single url spine, or up to 20 parts[], never both; the hint is use_audio_url_or_audio_parts.
Sume timeline audio_url_required: a silent Short needs mode silence
audio_url_required means the timeline has no audio.url, no audio.parts and no mode. Add a spine, or set audio.mode to silence with a duration_seconds.
compose 400: stack keys and overlay keys mixed in one layout
compose layout keys are stack (split, image_region, ratio) or overlay (position, width_ratio, margin_ratio). Mixing them is a 400. Match layout to operation.
Timeline slot under 0.2 s refused: fast-cut hooks for Shorts in Sume
Sume Timeline 1.0 needs every video slot duration to be at least 0.2 s. Merge sub-0.2 s flashes into a neighbor, or hold a still.
video-inspect 400: frames at and fps together on a Shorts hook
Send frames.at or frames.fps to Sume video-inspect, never both: both returns 400 video_inspect_frames_program_conflict. Pick at[] for a Shorts hook check.
video-inspect frames needs at or fps: probe a Short with frames:false
video_inspect_frames_program_required means frames is an object with no at or fps. Omit frames for 8 stills, or pass frames:false to only probe a Short upload.
Dim a bright product clip before captions: free check, then encode
Make white captions readable on a bright product clip: the free /v1/video-filter/check, then a dim op at the $0.02 encode rate, then Timeline.
Product spec sheet to video: Wan 3.0 footage, exact specs via compose
Make a spec-sheet video without a model re-typing your numbers: generate the footage with Wan 3.0, then put your own spec card on screen with Timeline compose.
Square video to 9:16: crop 0.5625 of the width, or fill with blur?
A 1:1 clip becomes 9:16 by cropping 0.5625 of its width (x 0.2188) or by letting Timeline fit it with blur. How to choose, with the numbers.
TikTok video width and height: 1080x1920, and how to export it
TikTok width and height as third-party guides state them, 1080x1920 at 9:16, and two Sume routes to that file: crop and scale, or a Timeline render.
3:2 camera clip to 9:16: crop width 0.375, scale to 1080x1920
A 3:2 clip needs a crop of 0.375 of its width, centered at x 0.3125, to become 9:16. The video-filter request, the rounding, and what you lose.