Blog
Sume blog: agent-native video generation, Avatar 1.0 orchestration, and production APIs for builders and coding agents.
Trim 12 Shorts from one recording with asyncio and a semaphore
Cut 12 Shorts from one recording with asyncio and httpx: 12 trims at $0.02 is $0.24. A semaphore of 4 caps in-flight requests; the limit is your choice.
Pause cuts from STT words: a 0.35 s gap rule and the 0.2 s trim floor
Turn STT word times into keep-ranges: cut gaps over 0.35 s, drop slivers under the 0.2 s trim minimum, and price 8 trim jobs at $0.02 each.
Inspect transcript words without start or end: skip them in cuts
Inspect transcript words can omit start or end. Treat an untimed word as unknown, and never cut across a gap that contains one: 0 of its time is safe to remove.
Read probe width, height and rotation before 9:16 crop fractions
A 9:16 crop needs width = 9/16 x H / W: 0.3164 for 1920x1080 but 0.5625 for a square clip. Read the probe first, and swap W and H if it reports rotation.
Hide jump cuts with alternating punch-ins: one crop copy, N slots
One $0.02 video-filter crop makes a 1.25x punched-in copy. Alternate wide and punched slots at each pause cut for $0.02 total, not $0.02 per cut.
Reframe 16:9 to 9:16 before you burn captions, not after
Crop a 16:9 clip to 9:16 with video-filter (width 0.3164, x 0.3418), then caption. Order trim, crop, caption: $0.02 + $0.02 + $0.20 = $0.24.
Fit a voiceover to 30 seconds: tune TTS speed between 0.6 and 1.5
generation_config.speed runs 0.6 to 1.5 on Sume TTS. Three takes of a 450-character script cost $0.064125 total, so measure the file and retune.
Six podcast ad-read variants at 500 characters: TTS cost and tags
Six 500-character ad reads cost $0.1425 on Sume TTS ($0.02375 each). Tag each take with the metadata field and one Idempotency-Key per variant.
Pick the Sume video model in code: seconds, ratio, edit or swap
A 12-line Python function turns duration, aspect ratio, edit and swap needs into the models that can run the job. A 4 s 1:1 clip has none; a 12 s one has two.
Character turnaround sheet on a 4:1 strip, sliced into references
One 4:1 Nano Banana 2.1 image on Sume as a four-view character sheet at $0.15 (2K), sliced into four squares and fed back as input_references. Steps and cost.
The resolution enum has 8 values, but each Sume model takes 3 or 4
The /v1/videos resolution enum lists 360p to 4K. Wan, Seedance, H3, H3 Max and Omni each take a subset. Check supported_resolutions and what each extreme costs.
Does 9:16 cost more than 16:9 on Seedance 2.5? Same price, tested
On Sume, Seedance 2.5 costs $17.33 for 30 s at 720p in 9:16, 16:9 or 1:1. 21:9 is $17.42 and 3:4 is $17.44, so a vertical Short pays no aspect premium.
How many 480p drafts cost one 1080p render? Wan 3.0 and Seedance 2.5
A 9-second Wan 3.0 clip is 4 drafts to 1 final at 1080p; Seedance 2.5 is 5.29 drafts per 1080p. The break-even table, priced from Sume rates.
Build the pricing_skus key from the resolution string: 4K is uppercase
Omni Flash 1.1 rates sit at per-video-second-360p, -720p, -1080p and -4K. Build the key from a template, mind the 4K case, total 10 s from $0.375 to $3.75.
Thousand-clip batch: the cents a frame ratio adds on Seedance 2.5
On Sume, 4:3 costs 1.7% less than 16:9 at 480p but 0.6% more at 720p and 1080p for a 10-second Seedance 2.5 clip. 3:4 mirrors it against 9:16.
Omni Flash 1.1 takes no audio reference: where a Short's voice goes
Gemini Omni Flash 1.1 accepts image and video references but not audio_url. For a voiced Short use native audio or a Timeline spine; 60 s is $7.60 at 720p.
A 1,850-character narration in Sume TTS: price, and what 3 takes add
A 1,850-character script costs $0.0879 in Sume TTS at $0.0475 per 1,000 characters; three takes cost $0.2636. Table by length, and how characters count.
Per-request timeout on Sume video polls: 15 s abort, then poll again
Give every poll GET its own 15 s AbortSignal.timeout so one hung read does not freeze a 30 s Wan or Seedance job loop. Reads per job per hour at 5, 10, 30 s.
H3 Max 4K request fails: H3 takes 2K and 4K as 768p upscales
minimax-h3-max accepts 480p, 768p or 1080p only. A 4K request belongs on minimax-h3, where it is a 768p upscale at $0.20 a second, $1.60 for 8 s on Sume.
TTS 1.0 rejects a model field: use the TTS router to pick sonic-3.6
Sume TTS 1.0 does not accept model. To choose an engine such as sonic-3.6 use /v1/tts-router/generate. Price is the same $0.0475 per 1,000 characters.
A 60-second video from six 10-second clips: H3, Wan 3.0, Seedance
Six 10-second clips make a minute: $4.50 on MiniMax H3, $7.50 on Wan 3.0, $22.68 on Seedance 2 and $34.668 on Seedance 2.5, before retries.
9-second vertical story ad on Seedance 2.5, and the marginal second
A 9-second 9:16 Seedance 2.5 clip costs $2.42 at 480p, $5.20 at 720p and $12.79 at 1080p on Sume.
CI guard: fail the build when a new Sume pricing_skus key appears
A node:test file fetches GET /v1/videos/models and fails when a model uses a pricing_skus key your cost code cannot total. Wan, Omni and H3 rates in the table.
Narrate a listing in your avatar's voice: TTS with avatar_handle
Sume TTS 1.0 can speak with the voice of a ready avatar if you pass avatar_handle. A 1,200-character listing script costs $0.057.
MCP generate_video max_spend_usd for Wan 3.0: 480p, 720p, 1080p
What to put in max_spend_usd when an MCP client calls generate_video for Wan 3.0: $0.625 to $7.50 for 10 and 30 seconds. Includes the dry_run body.
Recast 480p fails: resolution must be 768p or 1080p, priced by length
H3 Max Recast refuses 480p with 'resolution must be 768p or 1080p'. A 10-second source costs $3.75 at 768p and $5.625 at 1080p on Sume. The 5 to 30 s table.
Six Nano Banana 2.1 start frames, six Wan 3.0 clips: a 3-min Short
Six 9:16 Nano Banana 2.1 stills ($0.60 at 1K) feed six 30 s Wan 3.0 jobs as first frames: $23.40 at 720p for a 180-second Short, with Timeline.
usage.cost on a Sume video poll is nullable: guard it before you sum
The /v1/videos poll schema allows usage.cost to be null. Sum only completed jobs, keep unknowns apart in a ledger, and compare to the Wan 3.0 $3.75 estimate.
1:8 vertical strip split into four 1:2 panels: cost vs four calls
A 1:8 Nano Banana 2.1 strip on Sume cuts into four 1:2 panels for a long scroll. $0.15 at 2K vs four calls on models that list 1:2. Math and a request.
Agent Completion: a 15 s MiniMax H3 clip plus voiceover, cap 1.50
A 15-second MiniMax H3 clip at 768p ($1.125) plus a 750-character voiceover ($0.035625) comes to $1.160625, so a cap of 1.5 is enough. Table and body.
Reconcile a 9-clip batch: poll usage.cost versus /v1/usage debited_usd
If 7 of 9 Seedance 2.5 clips complete at $3.4668 each, the poll sum and the ledger should both read $24.2676. How to compare them and why they differ.
8:1 website section divider at 1440x180: cost of 5 dividers on Sume
A 1440x180 divider is 8:1, a ratio Nano Banana 2.1 lists on Sume. Five dividers cost $0.50 at 1K, $0.75 at 2K. CSS aspect-ratio and a request that matches.
Omni Flash rejects 768p: the resolution strings each video model takes
Omni takes 360p, 720p, 1080p or 4K; H3 takes 768p, not 720p. One table of the accepted resolution strings per model and the price range of each on Sume.
Re-render 3 of 17 rejected clips: what the retries add on Seedance 2
Seventeen 8-second Seedance 2 clips cost $51.408; re-rolling 3 rejected clips and then 1 more adds $12.096. Failed jobs are released, rejected takes are billed.
Podcast video episode over 30 minutes: Sume's 1800-second source limit
Sume video trim, audio detach and video inspect all cap a source clip at 1,800 seconds. A 45-minute podcast video must be split before you import it.
Ultra-wide title cards: five 6 s Seedance 2.5 clips in 21:9 at 1080p
A 6-second 21:9 Seedance 2.5 clip at 1080p costs $8.57 on Sume, about 4 cents over 16:9. Table by ratio.
pricing_skus has three key shapes: one Python function for the total
GET /v1/videos/models returns per-video-second-<res>, per-video-second and per-1000-video-tokens. A Decimal function totals a 30 s clip and flags token pricing.
320x50 and 468x60 banners from one 8:1 image: keep 80% and 97.5%
A 320x50 slot is 6.4:1 and 468x60 is 7.8:1. Generate one 8:1 Nano Banana 2.1 image on Sume for $0.10 at 1K, then crop. The width you keep and what to avoid.
TypeScript: check 8:1 against /v1/images/models before you POST
A 20-line TypeScript guard that reads the aspect_ratio values for a Sume image model and refuses a ratio it does not list, instead of a 400 at request time.
Nano Banana 2.1 edit vs generate on Sume: same price per image by tier
Do reference images change the Nano Banana 2.1 price on Sume? The estimator reads tier and image count only: $0.075 to $0.20 per image, edit or not.
Webhook and poll race on a Sume video: one SQLite insert decides
Keep polling as a backup to the job webhook. Dedupe on job_id with INSERT OR IGNORE so only one path downloads. 10 deliveries 30 s apart cover 270 s.
Four-week video budget: Omni hooks, H3 b-roll, two Recast swaps
A four-week plan with 3 Omni hooks and 2 H3 b-roll clips a week plus two 15-second Recast swaps totals $29.25 on Sume. Every line is priced.
Alert when a pricing_skus rate moves: a nightly diff of the model list
Save GET /v1/videos/models rates each night and diff them. A one-cent-per-second move costs $0.30 on a 30 s clip and $300 over 1,000 of them. Python, 25 lines.
How many jobs make a 3-minute vertical video: 6, 12 or 18 by model
A 180 s vertical video needs 6 jobs on Wan 3.0 or Seedance 2.5, 12 on MiniMax H3 and 18 on Omni Flash 1.1, since each model has a different maximum clip length.
Compare GET /v1/balance micros with a clip total, in integers
Balance comes as USD micros and cents. A 30 s Seedance 720p clip is 17,334,000 micros, 1,733 cents rounded down. Integer compare, plus the expiring-soon fields.
Contract test for 1:4, 4:1, 1:8, 8:1: pytest on /v1/images/models
A 12-line pytest that fails CI if Sume stops listing the Nano Banana 2.1 strip ratios. It reads GET /v1/images/models, so it generates nothing.
Job opening video with an AI host: 16:9 and 9:16 versions
A 30-second hiring video with a Sume avatar costs $5.52 on Standard or $7.35 on Plus per aspect ratio. A 16:9 and a 9:16 cut on Plus is $14.70.
Support answer video with an avatar: 25 seconds with a product shot
A 25-second support answer video costs $4.60 on Standard, $6.125 on Plus and $13.75 on Max. A product image adds a small per-second premium.