Blog
Sume blog: agent-native video generation, Avatar 1.0 orchestration, and production APIs for builders and coding agents.
Veo 3.1 Standard, Fast and Lite per-second prices vs Sume rows
Google lists Veo 3.1 at $0.05 to $0.60 per second. Sume's catalog has no Veo row, so this compares Google's prices with Wan 3.0, Omni Flash and Kling on Sume.
Vercel 800 s max duration: do you still need a Sume webhook?
Vercel Pro allows 800 s functions and a 30-minute beta. A Sume video job can still outlast one request, so use async or webhook mode and return in seconds.
Walmart 100MB vs Amazon Sponsored Brands 500MB: bitrate each allows
Walmart caps video at 100MB, Amazon Sponsored Brands at 500MB. The average bitrate each allows for 15 to 90 seconds, plus the Sponsored Brands 1 Mbps floor.
Webhook signature mismatch: Sume fingerprint vs OpenRouter t=,v1=
A video webhook that fails verification has three usual causes: parsed body, wrong secret, wrong header format. Sume adds a secret fingerprint header.
Test a video webhook receiver before the first job: Sume vs fal
Sume sends a signed dummy webhook.test with one API call. fal retries real results up to 31 times and treats 3xx as failure, so test your URL first.
Wedding save-the-date video with Omni: 8 seconds, vertical
A 9:16 save-the-date from an engagement photo with Gemini Omni Flash: $1.00 at 720p on Sume, with names and date added as captions, not generated text.
Which video model for a 12 s 9:16 clip with end frame and sound?
Seedance 2.5, Wan 3.0, Kling 3 and MiniMax H3 all take a 12-second 9:16 clip with an end frame on Sume. Prices run from $0.90 to $6.94; here is the pick.
Wine tasting invite video: Wan 3.0, 9:16, 12 s, audio reference
A 12-second vertical invite on Wan 3.0 costs $0.75 at 480p and $1.50 at 720p on Sume, and takes up to 5 audio references. Age gates and limits.
Yoga class promo video: three beats in one 10-second clip
A studio promo in 9:16 with Gemini Omni Flash: three timecoded beats in 10 seconds, calm native sound in the prompt, $1.25 at 720p on Sume.
YouTube AI disclosure: a clip of someone giving advice they never gave
YouTube's disclosure page lists making someone appear to give advice they never gave. Its examples in a table, and what Sume trims and captions change.
Shorts ads: Google says sound lifts conversions 20%+, add a bed
Google's Shorts ad page says sound lifts conversions by over 20%. Add a music bed that ducks under a voiceover with the Sume Timeline soundtrack field.
Zoom deepfake risk detection: what it means for avatar video
Zoom lists real-time alerts for synthetic audio or video next to meeting avatars. What the page says, what it leaves open, and how to label a Sume clip.
MiniMax H3 Max 1080p costs twice 768p: the three tiers priced
H3 Max (minimax-h3-max) on Sume: 480p $0.0625, 768p $0.10, 1080p $0.20 per second, 5 to 15 seconds. 1080p is refined from native 768p.
Fix wrong text on a product label with ideogram-v4.5 and references
ideogram-v4.5 edits the first image and takes up to four more as references, five in total. Send a label photo plus a type sample; cost per edit by quality.
Extract 16 kHz mono audio from a video for transcription
A Python script that detaches speech-ready 16 kHz mono wav from a Sume-hosted video, then polls the job. Costs $0.01 per detach before the STT minute.
Assemble a three-clip montage with fades through the Sume API
A working Timeline 1.0 request that joins three hosted clips over one audio spine with fade and dissolve transitions, plus the Python to poll it, for $0.10.
Go: net/http client for a 30-second Wan 3.0 clip, 30 lines
A 30-line Go program using only the standard library: submit wan-3.0 for 30 seconds, poll the job, write clip.mp4. Reserve and per-second rate included.
Omni Flash 1.1 price chart: 3 to 10 seconds at 4 resolutions
Every whole-second length from 3 to 10 seconds at 360p, 720p, 1080p and 4K for Gemini Omni Flash 1.1 on Sume, from the estimator that reserves your balance.
Dim a background clip under text: video filter dim amounts
The Sume video-filter dim op multiplies luma by an amount between 0 and 1; black stays black and chroma is untouched. Amount table, request and the $0.02 price.
Seedance 2.5 1080p is 2.46x the cost of 720p: why
Seedance 2.5 1080p costs about two and a half times 720p per second on Sume. The pixel count, the higher 1080p token rate, and the other Seedance tiers.
Trim and conform to 1080x1920 at 30 fps in one video-trim job
Video trim's output field re-encodes to a set width, height and 24/25/30/60 fps in the same $0.02 job, exact precision only. Request, ranges and refusal codes.
Cheapest image model per 1,000 images: 18 Sume rows ranked
Higgsfield Soul is $5 per 1,000 images on Sume, gpt-image-2.5 low at 1024 is $7.38, and Nano Banana Pro is $187.50. Eighteen rows ranked with one request.
Sume MCP avatar tools: 5 read-only and the paid create set
Hosted Sume MCP lists five read tools for avatars and avatar videos and a paid group for creates and previews. What OAuth mcp:read sees and the dry_run flow.
Pre-flight an Omni Flash request against the Sume model catalog
A short Python check that reads supported durations, resolutions and aspect ratios for gemini-omni-flash-1.1 via /v1/videos/models.
Video upscale hold: omit duration_seconds and Sume reserves 5 s
Sume video upscale reserves from duration_seconds, or 5 seconds when you omit it. Hold table for 5, 15 and 30 seconds at $0.009 per second and the 402 case.
Dart: submit a 30-second Wan 3.0 clip with package:http
Server-side Dart with package:http: POST wan-3.0 for 30 seconds, poll until completed and write the MP4. Keep the key out of the Flutter app.
flux-2-pro vs flux-2-flex on Sume: price, limits and a 100-image bill
flux-2-pro is $0.0375 and flux-2-flex is $0.0625 per image on Sume. Both take 10 references and n up to 4. Compare the 100 and 1,000 image bill.
One 30 s Seedance 2.5 job vs two 15 s Mini jobs: cost and trade-offs
A 30-second Seedance 2.5 clip costs more than two 15-second Mini clips on Sume. The numbers at 480p, 720p and 1080p, and what two jobs cost you in continuity.
Join voiceover takes into one gapless wav: timeline audio concat
Timeline audio concat joins up to 20 hosted audio parts sample-exact into one reusable wav for $0.01 per job and returns segment offsets.
Avatar video for YouTube Shorts: 9:16 or 1:1, up to 60 seconds
Avatar 1.0 makes 9:16 and 1:1 talking videos of up to 60 seconds at 720p, inside the Shorts window. Cost for 30 and 60 seconds, and one request.
Video trim caps: 1800 s in, 900 s out, and the 901 s refusal
Sume video-trim accepts a source up to 1800 seconds and an output up to 900 seconds. The refusal codes you get past either cap and how to split a long cut.
Etsy-style listing gallery: 10 images per product, cost by image model
Ten gallery images for one product cost $0.07 to $1.00 on Sume depending on the model; 25 products is 250 images. Table with seedream, flux and gpt-image-2.5.
Seedance 2.5 vs Seedance 2: the price gap at 480p, 720p and 1080p
Seedance 2.5 costs 1.53x Seedance 2 at 720p and 1.67x at 1080p on Sume. Per-clip prices for 6, 10 and 15 s, and what 2.5 adds.
Budgeting 100 Omni Flash clips with a mixed-resolution plan
A 100-clip Gemini Omni Flash plan across four resolutions costs $56.125 on Sume, versus $150.00 if every clip were an 8-second 1080p render.
Face swap request in Python: validate the video URL before you send
A Python snippet that rejects non-HTTPS, localhost and private-IP video URLs locally, then submits a Sume face-swap beta run with a quality and idempotency key.
Music API 400s: duration and negative_prompt, and the fix
Sume's Music Router returns 400 if you send duration, duration_seconds or a non-empty negative_prompt. The fix for each, and a request that passes validation.
rate_limited or queue_full? One Python submit handler for both 429s
Two different 429s need two different waits. A Python handler reads error.code, sleeps on retry-after for rate_limited, and waits for capacity on queue_full.
Port an OpenRouter video client to Sume: base URL, key, model ids
Sume's /v1/videos follows the OpenRouter video wire. Three edits move a client: base URL, API key, bare model id. The six differences that still bite.
Balance needed to submit 10 or 50 video jobs: reserve per model
Sume reserves each job's estimate at submit. A table of the balance 10 and 50 ten-second clips need on six video models, and where a 402 lands in a batch.
Which Sume API errors to retry and which to stop on: Node wrapper
A retry policy by error.code for Sume submits: retry rate_limited, queue_full, provider_capacity_exceeded; stop on 400, 401, 402, 409. Node wrapper with jitter.
Why input_references fails on Imagen, Recraft, Soul and Qwen Max
Five Sume image rows list zero input_references: imagen-4 fast and ultra, recraft-v4, Soul and qwen-image-max. Check the catalog in Python first.
AI dubbing workflow on Sume: detach, transcribe, TTS, timeline
Sume has no one-call dubbing endpoint. Chain detach, STT, your translation, TTS and a timeline render for a 60-second video at about $0.16 plus translation.
Switch video models by changing one string: what can still break
On Sume's /v1/videos you swap the model id and keep the body. Duration range, resolution and aspect ratio are the three fields that may need adjusting.
Reserve, capture, refund: a 20-clip batch with 3 failures, 2 cancels
Worked example of how a Sume balance moves across a 20-job batch when 3 jobs fail and 2 are canceled while queued: what is held, captured and released.
Can a Sume avatar speak my own recording? Voice types explained
Avatar 1.0 talking video takes a script or per-scene voice blocks of type text or silence. What the docs list, what they do not, and how to plan a pause.
Remove backgrounds from 500 product photos: RMBG API cost and code
Sume RMBG 1.0 removes a background for $0.0225 per image regardless of size. Batch cost for 50 to 5,000 photos, the request body and a Python loop.
Sume webhook retries: 10 attempts, 30 s apart, 10 s timeout each
The delivery schedule for Sume job webhooks: 10 attempts, fixed 30 s spacing, 10 s timeout, about 4.5 minutes of retries, then redeliver and the status poll.
Retry a timed-out Omni Flash submit without paying twice (Node)
Node 18+ code that retries a Gemini Omni Flash 1.1 POST to Sume's /v1/videos on timeout or 429 with one Idempotency-Key, so a retry returns the same job.