Blog
Sume blog: agent-native video generation, Avatar 1.0 orchestration, and production APIs for builders and coding agents.
A zod schema for the Sume /v1/videos poll response, ported from Sora
Replace Sora's four-status type with a zod schema for Sume's five statuses. It rejects a completed job without unsigned_urls and a failed one without an error.
TikTok AI literacy hub: what your AI video should show viewers
TikTok's in-app hub teaches viewers to spot AI content when they search AI terms. What that means for a brand's AI video, and how to prepare yours with Sume.
YouTube Shorts thumbnail file type and size: JPG or PNG, 9:16, 50 MB
YouTube's help page gives JPG or PNG, 9:16, 2160x3840 and 50 MB on desktop for Shorts thumbnails. Export that from a video frame with Sume video frames.
Sume SDK 0.2.0 exported constants: webhook headers and wait defaults
@sume-com/sdk 0.2.0 exports the webhook header names, the 300 s tolerance and the wait defaults as constants. List them so your code does not repeat strings.
Instagram Series for Reels: prepare AI episodes before it opens
Meta is testing Series, a profile tab that bundles episodic Reels. Render numbered 9:16 episodes now with Sume so a season is ready to group when you get it.
Does each Shorts series episode need to stand alone?
Each Shorts episode may be the first one a viewer sees. Add a recap overlay with Sume trim, compose and Timeline 1.0 for about $0.15.
Higgsfield API vs Sume Videos API for a Sora port: five changes
Higgsfield's Sora migration guide changes auth, endpoints, body, response id and webhooks. What each costs in code, next to the Sume /v1/videos shape.
Prepaid or postpaid AI video API: how Sume bills, and invoices
Sume bills self-serve usage from a prepaid USD balance reserved at submit; Enterprise lists invoice and PO billing. A 30-second Wan 3.0 hold, worked.
Grok Imagine Video 1.5 API in Python: animate a still on Sume
Animate one image with Grok Imagine Video 1.5 on Sume: Python, a first frame, 720p, 4 to 15 seconds, $0.15 for 12 seconds. Text-only calls fail.
Charging users for AI video: price each one from usage.cost
After Sora, your app pays Sume's usage.cost per video. Add a margin, round up to the cent, charge only completed jobs. A JavaScript helper and rules.
Gemini Omni Flash 1.1 limits for a Sora remix: 10 images, 3 videos
gemini-omni-flash-1.1 on Sume takes 10 reference images, 3 videos of 3 seconds each, and a 20000 character prompt. Tokens like <IMAGE_REF_0> point at inputs.
Sora 2 vs Wan 3.0 after the API shutdown: length, size and price
Sora 2 and its Videos API ended 2026-09-24. Wan 3.0 on Sume does 2 to 30 seconds at 480p to 1080p; limits and billed price per second, side by side.
Wan 3.0 reference limits on Sume: 10 images, 5 videos, 5 audios
wan-3.0 accepts up to 10 reference images, 5 videos (15 seconds total, 16 fps or more) and 5 audios (15 seconds total). What to trim when porting Sora inputs.
Sora's GET /videos and DELETE /videos/{id}: what Sume offers instead
Sora had list and delete routes. Sume lists your key's own jobs with GET /v1/jobs and documents no public delete, so plan retention in your own storage.
Runway Aleph 2.0 vs Gemini Omni Flash edit as a Sora remix swap
Sora's remix endpoint is gone. Runway lists Aleph 2.0 video-to-video at $0.28 per second; Sume's Gemini Omni Flash edit takes video_url at $0.125 per second.
Save a sidecar JSON with every Sume mp4: job id, model, cost, prompt
Export finished video jobs with a Python script that downloads each mp4 and writes a sidecar JSON with job id, model, cost and the prompt you saved.
Re-render a prompts.txt of old Sora prompts with curl, jq and xargs
Turn a file of saved Sora prompts into Sume video jobs with four parallel curl calls, an idempotency key per prompt, then poll jobs.tsv and download each mp4.
Smallest AI video clip you can buy on Sume: shortest length, price
The shortest clips on Sume: Wan 3.0 from 2 s, Gemini Omni Flash from 3 s, Seedance from 4 s. Cheapest test render costs $0.1125, before a $0.0062 fee.
What does a daily 30-second AI short cost per month? Wan vs Seedance
Posting one 30-second AI short a day for 30 days costs $56.25 on Wan 3.0 at 480p and $520.02 on Seedance 2.5 at 720p in video alone, before the 5.5% fee.
30-episode avatar series cost: Sume Avatar standard, plus and max
A 30-episode talking-avatar series of 45-second episodes costs $248 to $743 in avatar video on Sume, plus $0.95 to create the avatar. Per-second math inside.
Thumbnails for a 30-episode series: 3 options per episode, priced
Generating 3 thumbnail options for each of 30 episodes is 90 images. On Sume that costs about $0.47 to $9.50 across eight image models, with the fee. Worked.
Cost of one finished minute of AI video: video, voice, music, timeline
One finished 60-second video on Sume costs about $4.23 to $90.26 depending on model and resolution, with voice, music, timeline and fee included. Worked.
Finishing a 30-episode series: captions, upscale, cutouts on Sume
Captions, cutouts and upscales for 30 episodes of 45 seconds: speech-to-text is $0.225, video upscale $12.15. Computed from Sume rates.
30-episode AI Shorts series budget: 45-second episodes, priced
What a 30-episode series of 45-second AI episodes costs on Sume: video, voiceover, music and timeline per episode, then x30 plus the 5.5% fee, worked.
Runway Muse Image at 1 credit vs Sume image models: price per 1,000
Runway lists Muse Image at 1 credit, $0.01 each, or $10 per 1,000. Sume's cheapest image model is $5 per 1,000 and the dearest of eight is $100, before fee.
Threads video over 5 minutes: split an AI video with Sume video trim
Threads takes videos up to 5 minutes. How to cut a longer AI video into parts with POST /v1/video-trim, and what each cut costs on Sume.
Trim, caption, compose: Sume returns new files, keeping your original
Sume's trim and caption tools return new MP4s and leave the source unchanged. Why that keeps the original AI generation intact, and what each step costs.
Meta's AI Info label hits identical copies: post one master AI video
Meta says an AI Info label applies to all identical instances across its platforms. Why one master file matters for AI video, and how to keep one with Sume.
Gemini Omni Flash 1.1 price: Runway vs Google vs Sume per second
Gemini Omni Flash video costs about $0.10 a second at 720p on Google's page and 10 credits ($0.10) on Runway. Sume sells it at $0.125 before the 5.5% fee.
Price per second of 720p AI video on Sume: Wan, Omni, Seedance ranked
Ranked 720p price per second on Sume: Wan 3.0 and Gemini Omni Flash 1.1 at $0.125, Seedance 2.0 mini $0.1890 up to Seedance 2.5 at $0.5778, before the fee.
Veo 3.1 Standard, Fast, Lite: price per 8-second clip vs Sume
An 8-second Veo 3.1 clip costs $3.20 Standard, $0.80 Fast and $0.40 Lite at 720p on Google's page. Wan 3.0 and Omni Flash on Sume are $1.25 before the fee.
YouTube Creator Partnerships brand Short: what format to deliver
What file a brand needs for a YouTube Creator Partnerships Short: length, ratio and how to cut it with Sume.
Cut 15 s and 30 s versions of a creator Short for a brand partner
A brand may ask for shorter cuts of a creator video. Make 15 and 30 second versions from one file with Timeline 1.0 source_in, at $0.10 per ceil minute each.
Paid promotion on a branded Short: YouTube box plus on-screen line
Tick YouTube's paid promotion box, then add your own opening line with Sume trim, compose and Timeline 1.0 for about $0.15.
Omni Flash edit or regenerate: when video_url beats a new prompt
Edit with video_url to change one thing; regenerate when the motion is wrong. A 6-second edit costs $0.75 at 720p on Sume.
Which Sume video models take a source video_url? Edit, transfer
Only three Video Router rows accept a video_url source: gemini-omni-flash-1.1 (edit), higgsfield-genjutsu (motion transfer) and h3-max-recast (person swap).
Sign a test job.completed webhook and replay it at your receiver
Build a Sume-style job.completed delivery in Python: HMAC over timestamp.body, a verifier that refuses an empty secret, plus tamper, stale and rotation checks.
OpenAI webhook retries (72 hours) vs Sume (10 attempts): what changes
OpenAI retries a failed webhook for up to 72 hours; Sume tries 10 times, 30 seconds apart. Rebuild the receiver around an inbox, job-id dedupe and polling.
Sume TTS cost by length: 1 to 20,000 characters, rounded up
Sume TTS bills $0.0475 per 1,000 characters and rounds each job up to the cent. A ladder from 1 to 20,000 characters shows where the rounding costs you.
Sume TTS tts_duration_exceeded: splitting a script over 1,200 seconds
Sume TTS fails audio longer than 1,200 seconds with tts_duration_exceeded, and caps a request at 20,000 characters. Split at paragraph breaks and rejoin.
Seedance 2.5 price by aspect ratio: 9:16 vs 16:9 vs 1:1 on Sume
Does a vertical 9:16 Seedance 2.5 clip cost more than 16:9? On Sume a 10-second clip costs the same; 1:1 at 480p is cheaper, 4:3 slightly dearer. Table inside.
Seedance 2.0 mini, fast, standard vs 2.5: price of a 15-second clip
A 15-second 720p clip on Sume costs $2.84 on Seedance 2.0 mini, $4.54 on fast, $5.67 on standard and $8.67 on 2.5, before the 5.5% fee. Arithmetic shown.
Language listening practice: one passage at 0.6, 1.0 and 1.5 speed
Make slow, normal and fast versions of one passage with Sume TTS generation_config.speed (0.6 to 1.5). Three jobs for 9 cents on a 600-character text.
Sume TTS 409 tts_voice_language_mismatch: confirm and retry safely
A Sume TTS 409 means the voice's language differs from your request. Nothing was charged. Ask the user, then retry with confirm_language_mismatch true.
Ask Studio says my Short drags: cut the dead air in Python
When feedback says a Short drags, find pauses from word timings and rebuild it without them. Tested Python and the Timeline 1.0 body.
Announce a new Shorts episode with a YouTube post and a video frame
Make the still for a YouTube post that points to your next Shorts episode: pull the frame with Sume video frames and keep within the post limits YouTube lists.
Gemini 3.8 Flash TTS vs Flash-Lite TTS: what one hour of audio costs
One hour of speech is 90,000 audio tokens. Flash TTS is $0.81 and Flash-Lite TTS $0.54 at the 2026 standard rate, doubling in 2027. All tiers compared.
Claude Code 2.1.290 claude -p skips resources/list: Sume is tools-only
Claude Code 2.1.290 no longer holds the first claude -p turn for HTTP MCP servers to answer resources/list. Sume declares tools only, so nothing is lost.