Use cases
Worked examples on Sume's catalog Formats and endpoints: product video and photography, try-on, app and product ads, logo animation, and more.
Start with Ready-made Formats for product video: the Sume Format catalog
A/B test two voices on one ad script: cost of two takes
Voice both takes of a 1,000-character ad with two Sume voices for $0.095. Compare length and sound, then render only the winner for $0.10 a minute.
Coffee or tea pouch photos with AI: zip top, crimp and print
Pouches have a crimped top, a zip and a window. Stage them on a counter with Sume's image API and check the seal, the print and the shape before publishing.
Supplement bottle ad photo with a legible label: Pillow composite
AI scenes garble small label print. Generate the bottle scene with Sume's image API, then paste the real label crop back with Pillow so claims stay exact.
H3 Max Recast in paid ads: what fal says on commercial use
fal's H3 Max Recast page lists commercial use with no licence text. What it covers, and the releases and source rights you still need before a it runs as an ad.
Pre-render a voice line bank for stream alerts instead of live TTS
Alerts, kiosks and game barks repeat a few lines. Render each once with Sume TTS and play the stored file. Thirty 80-character lines cost about 11 cents.
Swap the person, then the product: Recast and Omni edit in turn
One ad clip, two changes: Recast replaces the presenter from a photo, then Gemini Omni Flash 1.1 swaps the product by prompt. Two jobs, two prices.
Watch or bracelet try-on from a wrist photo: two references, one call
Show a watch or bracelet on a shopper's own wrist with Sume's image API: wrist photo plus product photo as two references. Request shape, checks and cost notes.
Watch product photography with AI: dial, glass reflections, case size
Watches need exact dials. Make lifestyle watch photos with Sume's image API from one reference, and check hands, markers, crown and glare before publishing.
Annual compliance refresher voiceover in six languages: what it costs
A 3,000-character training script in six languages costs $0.855 on Sume TTS. Set language per voice, translate yourself, and re-run only what changed.
A 180-second Short from six Wan 3.0 clips: cost and Timeline join
YouTube allows Shorts up to 3 minutes. Six 30-second Wan 3.0 clips at 720p cost $22.50 on Sume, plus a $0.30 Timeline join, for one 9:16 video.
City-localised ad backdrops: one product photo, many cities
Run one product ad in ten cities by changing only the backdrop. A Sume image API loop with a fixed product prompt, per-city results and a cost log, with checks.
A 90-second science explainer from stills, narration and captions
Turn six diagram stills and a 1,200-character script into a captioned class explainer for about $0.66: speech at $0.057, a $0.10 render and $0.20 captions.
Stage one chair in five rooms: references, n and cost per room
Show one chair in five styled rooms with Sume's image API: one product reference, a room per prompt, and a way to keep the chair's shape stable.
Handbag product photos: hardware close-up as a second reference
Keep clasps, zips and stitching accurate in AI handbag photos: send the full bag and a hardware close-up as two references to Sume's image API. Checks included.
Hat and cap try-on API: one selfie, one call per colorway
Let shoppers see a cap or beanie on their own head: one selfie plus one product photo per colorway, a cache key per shopper, SKU and colour, and a cost log.
Headphone and earbud product photos: logo and orientation QC
Make headphone or earbud lifestyle shots with Sume's image API, then catch flipped logos, wrong earcup sides and cable errors with a Pillow contact sheet.
A 15-second home-anniversary video for past clients from one avatar
Send past buyers a short personal video each year. One $0.95 avatar and a 15-second script cost $2.76 on standard or $3.68 on plus per video.
Jacket try-on over the shopper's own outfit: prompt wording that holds
A jacket goes over clothes, not instead of them. Prompt wording and a two-reference Sume image call that keep the shopper's shirt visible, plus layering checks.
Kettle or air fryer product photos: cord, spout and counter scene
Kitchen appliances fail on cords, handles and dials. Generate counter scenes with Sume's image API from one reference and check the parts that go wrong first.
Lamp product photos, lit and unlit, from one reference
A lamp needs an off photo and an on photo. Use one reference and two Sume image calls with matched prompts, then check that the shade and base stay identical.
Turn LRC lyrics into caption cues for a lyric video (Python)
A 26-line Python script converts an LRC lyrics file into the cues array for POST /v1/video-captions, so a silent lyric video is captioned for $0.20.
One layout reference, twenty posters: GPT Image 2.5 on Sume
ChatGPT Images 2.5 ships poster templates. On the API, reuse one layout image as a reference and change only the headline: 20 calls.
Planter and plant pot photos with AI: scale, drainage and plant
Staging a planter with a plant in it changes how big the pot looks. Sume image API prompt and checks for scale, drainage hole, saucer and the plant itself.
A listing walkthrough in English and Spanish from one silent video
Narrate one silent walkthrough twice, with two text-to-speech tracks at $0.0475 per 1,000 characters, then join each with captions. About $0.35 a language.
Reformat a vertical Short to 1:1 and 16:9 with a blur fit
Timeline's blur fit fills a square or widescreen frame with a blurred copy of your 9:16 clip. Each reformat costs $0.10 per output minute on Sume.
A holiday-hours video for your Google Business Profile in 30 seconds
Google's help page allows Business Profile videos up to 30 seconds, 75 MB and 720p. Make a holiday-hours clip with an avatar for $3.88 or $5.16 with an image.
Restyle holiday ad captions for an A/B test without re-transcribing
Send source_caption_id to burn the same ad in a new caption style. Sume does not run speech-to-text again, and each restyle is a $0.20 render.
Try-on with front and back garment photos: three references
When a garment's back has a print the front lacks, send shopper, front and back photos as three references. Prompt wording, a Python call and what to check.
Tumbler or water bottle photos with AI: composite the logo in code
Brand logos on bottles drift in generated images. Make the scene with Sume's image API, then paste your transparent logo with Pillow so the mark is exact.
Write a 15-second holiday ad song with a timestamped Lyria brief
Sume Music has no duration field, so length goes in the prompt. A brief with [0:00-0:05] sections gets a 15-second holiday jingle for $0.125 a try.
Dub a 60-second clip into Hindi: Sume step costs before translation
A 60-second dub on Sume is four steps. Detach audio $0.01, STT $0.01, translate with your own tool, then TTS in Hindi: about $0.0628 for 900 characters.
Numbers, dates and prices in TTS: write the spoken form first
How to prepare a script for Sume TTS so numbers, dates, prices and acronyms are read as you intend, with a test sentence and a pronunciation dictionary.
Launch kit from 6 references: 4 angles, a style card, a logo
ChatGPT Image 2.5 on Sume takes up to 16 references, so one request can carry four product angles, a style card and a logo. How to order and prompt them.
Virtual staging 8 rooms, 3 takes each: Seedream vs Flux cost
A listing with 8 rooms and 3 staged takes each is 24 images: $1.20 on Seedream 4.5, $1.50 on Flux 2 Flex, $2.40 on Nano Banana 2 at 1K on Sume.
Animate a locally generated image with Sume video: first frame
Made a still with a local open-weights model? Host it at a public HTTPS URL, send it as first_frame to POST /v1/videos, and poll. Plus the licence check first.
Recast one video ad with several faces: consent checklist, cost
Reuse one 15 second ad with different people via H3 Max Recast on Sume: cost per variant, the release to get from each person, and how to label them.
Will a dubbed YouTube Short fit in 3 minutes? Measure before render
A dub in another language often runs longer than the source. YouTube Shorts allow up to 3 minutes; Sume does not cap you there, so check the length first.
Dub a video where two languages are spoken: keep the original lines
Sume's speech-to-text takes one language hint per call. For a mixed-language video, transcribe, translate only some lines, and splice the rest back in.
Can a brand make the video for a creator's Meta partnership ad?
Meta's branded content policy says creators cannot be paid to post content they were not involved in making. What that means for brand-made AI clips.
Affiliate link on Instagram: does it need the paid partnership label?
Meta says content with an affiliate link should carry the Paid partnership label, even with no business partner tagged. What that means for AI clips.
Partnership ad code or account-level permission: which to ask for
Meta partnership ads come in two permission types. How content-level codes and account-level access differ, and what to log when the clips are AI-made.
Pre-render voice agent greetings and fixed replies as TTS files
Live TTS latency is wasted on lines that never change. Render the greeting, hold message and confirmations once as files with Sume TTS and play them instantly.
Turn a creator's raw clip into a brand-avatar ad with face swap
A workflow for re-casting your own or a licensed creator clip onto a brand avatar: trim to 4-15 s, swap on Sume, caption. Python sample and costs by tier.
Edit one region of an AI image and keep every other pixel
A mask edit is guidance, not a guarantee. Paste the original back outside the mask with Pillow so untouched pixels stay identical. Python with Sume's mask_url.
Face swap on someone else's Short: does YouTube call it original?
YouTube said Oct 1 that Shorts re-uploading others' videos without significant changes will get less reach. Why a face swap on borrowed footage is a risk.
AI avatar of a real person: release checklist before the photo
Before you turn an employee's or creator's photo into a Sume avatar, get a written release. A checklist for scope, term and revocation, matched to the API.
Does TikTok keep Content Credentials on a downloaded AI video?
TikTok says it is working to attach Content Credentials that stay on download. What TikTok says, what its 2024 timeline promised, and how to test a file.
AI backing track generator: an instrumental to sing or play over
Generate an AI backing track by prompt: tempo, key, form, no vocals. Sume returns one mixed MP3, no stems or click track, so trim and check the take by ear.