Guides
Sume guides: the API, Formats, agents, models, media tools and pricing, each answer dated and sourced.

Gemini Omni video editing: the EEA, UK and under-10-second rules
Google says editing or extending uploaded videos with Omni is unavailable in the EEA, Switzerland and the UK, and uploads must be under 10 seconds.

Gemini Omni first and last frame: image_url + end_image_url
Google calls it Interpolation (first + last frame). On Sume you send image_url plus end_image_url to gemini-omni-flash-1.1 through the Video Router.

Gemini Omni extend video: Google's 40 s cap vs Sume's modes
Google's Gemini Omni can extend a video by 3-10 s up to 40 s. Sume's Omni row has no extend mode; here is what it offers instead.

Gemini TTS API: which engine Sume's TTS routes to
Gemini 3.8 TTS is not in Sume's TTS Router, which lists Cartesia Sonic ids only. What Sume's TTS takes for voice, engine and length, with a checklist.

Gemini TTS regional voices vs Sume's language tag
Google says Gemini 3.8 TTS covers Mexican Spanish and Quebec French. Sume's language takes a BCP-47 tag of 2 to 16 characters; dialect output is not promised.

Genjutsu Object Swap API: Sume serves Motion Transfer only
Higgsfield pairs Motion Transfer with Object Swap in Genjutsu. Sume's Genjutsu row serves Motion Transfer only; Object Swap is rejected by name.

Genjutsu 1080p: Sume serves it at 480p or 720p only
Higgsfield lists Genjutsu output up to 1080p. On Sume, Genjutsu Motion Transfer accepts 480p (default) or 720p, and any other value is rejected.

Genjutsu reference images: 30 on Higgsfield, 1 to 8 on Sume
Higgsfield says Genjutsu takes up to 30 reference photos. Sume's Genjutsu Motion Transfer row requires 1 to 8 images, so pick your best eight.

Genjutsu video length limit: 4 to 30 seconds, inspect first
Genjutsu takes a 4 to 30 second source video. On Sume the duration field must equal the inspected length, rounded up. Probe it with video inspect first.

GitHub Actions retention days: keep the Sume job id elsewhere
From 1 October 2026 GitHub Actions runs, checks and statuses follow the log retention setting (default 90 days). Store Sume job ids outside the log.

Google Ads AI label setting: a label inside the video
Google now allows an AI label inside AI video ad creatives without tripping text-overlay policy. Burn your own label line as the last step with Sume.

Google Ads election ads: the altered or synthetic checkbox
Election ads with synthetic or altered content need the Altered or synthetic content checkbox in Google Ads. Some formats get an automatic label.

Google Ads AI disclosure labels: add your own, away from the edges
Google says you can add your own AI label to an ad asset, keep it off the edges, and opt out of enhancements that crop it. Set margins with Sume compose.

Google Ads AI labels: SynthID and C2PA cover Google-made assets
Google Ads embeds SynthID and C2PA in assets made inside its own tools. An MP4 made elsewhere and uploaded relies on your AI label setting instead.

Asset Studio video in 16:9 and 9:16: two Sume requests
Google's Asset Studio makes horizontal and vertical video in one workflow. With Sume, send one brief as two requests, 16:9 and 9:16, with model limits.

Google Cloud Tasks retry per task: pair it with a Sume key
Cloud Tasks now sets retry parameters per task (GA 30 Sep 2026). When a task calls a Sume submit, send the same Idempotency-Key on every attempt.

Replace a Merchant Center product video: new filename and URL
Google says video_link URLs must stay stable and never be reused for another video. Each Sume trim or filter returns a new artifact URL, so a re-cut fits.

Google Merchant Center video link: 6 to 240 seconds, trimmed
Google Merchant Center video_link needs a video of 6 to 240 seconds. Probe the length, then cut a longer clip to size with Sume Video trim into a new MP4.

Merchant Center product video 720p minimum: what to request
Merchant Center video_link needs at least 720p (1280x720). Sume's generated video defaults to 720p; probe a file first, then conform with Video trim if needed.

Merchant Center video link aspect ratio: 9:16, 16:9, 1:1
Merchant Center video_link accepts only 9:16, 16:9 or 1:1, so a 4:5 clip is out. Sume Timeline and Video trim can output a size that fits one of the three.

Merchant Center "Video not processed": use a raw MP4 URL
Fix Merchant's "Video not processed" by submitting a URL to a raw video file, not a page with a player. A media.sume.com MP4 URL is a direct file link.

Google Play preview video length: the first 30 seconds
Play Console says only the first 30 seconds of a preview video autoplays. Trim a lead 30-second cut, and review its opening stills, with Sume.

gpt-4o-transcribe-diarize retiring: Sume STT has no speaker labels
OpenAI lists gpt-4o-transcribe-diarize for removal on Feb 26, 2027. Sume STT returns words and sentences with timings, but no speaker field.

Grok Imagine first and last frame: keyframes on xAI only
xAI takes up to 4 timestamped keyframes for Grok Imagine video. Sume's grok-imagine-video-1.5 row takes one start image and rejects an end frame.

Grok Imagine: how many images per request, xAI vs Sume
xAI's image API takes n from 1 to 10 per request. Sume's catalog code gives Grok a maximum of 1, so send one image per call and loop for more.

Grok Imagine video editing: 8.7 s on xAI, Omni edit on Sume
xAI caps Grok Imagine video edits at 8.7 seconds. Sume's Grok row has no video-to-video; edits go through gemini-omni-flash-1.1 with a video_url.

Grok Imagine video length: 1-15 s on xAI, 4-15 s on Sume
xAI allows 1 to 15 seconds for grok-imagine-video-1.5. Sume's catalog row allows 4 to 15, so a 2-second clip needs another id such as wan-3.0.

Grok Imagine video sound: xAI audio vs Sume's silent row
xAI's Grok Imagine video makes audio unless generate_audio is False. Sume's grok-imagine-video-1.5 row lists audio: false and rejects generate_audio.

Hy-Image-3.5-Preview API limits vs Sume's image limits
Tencent's Hy-Image-3.5-Preview takes 256 to 8192 px edges, up to 4096x4096 and 20 references. Sume lists no Hy Image model; its own limits are below.

Ideogram 4.5 Precise Edit API: mask and reference limits
Ideogram 4.5 Precise Edit takes 4 reference images, 3 when a mask is sent. Sume lists Ideogram V3, and mask_url there is only for ChatGPT Image 2.5.

India's synthetic-media label rule: does routine editing count?
India's amended IT rules exclude routine editing, formatting and compression from synthetic media, and require a prominent label on the rest. What Sume offers.

Inngest step return 4 MiB limit: return the Sume artifact URL
Inngest caps one step return at 4 MiB and run state at 32 MiB. Return the media.sume.com artifact URL from step.run instead of the video bytes.

Instagram Audio API for Reels ads: prepare the video side
Meta's Instagram Audio API can find royalty-free replacements for copyrighted Reels music. Sume does not call it, but can drop the audio from your video.

Instagram Graph API hashtag search: 30 per 7 days, and Sume
Meta caps hashtag search at 30 unique hashtags per account per rolling 7 days and needs app review. Sume's instagram_search_hashtag is a bounded public read.

Instagram oEmbed: embed HTML vs Sume's instagram_media read
Instagram oEmbed is meant only for embedding, at up to 1,000 requests per hour. Sume's instagram_media returns media candidates for a post or reel URL.

App Store preview video length: 15-30 s, 30 fps max
Apple's app preview specs list 15-30 seconds and a 30 fps maximum. Sume's Timeline keeps source fps when omitted, so set output fps to 30 for a preview.

Add Sume as an MCP server in Junie CLI (mcp.json)
Add Sume to Junie CLI's mcp.json as a remote server with url and headers, or leave headers out and use Authorize for OAuth. Both paths, step by step.

Kevin MacLeod attribution: credit text and Sume's BGM field
Kevin MacLeod tracks need a CC BY 4.0 credit that is easy to find. Sume's BGM featured beds are his tracks; keep track.attribution with any public use.

Kling 2.1 API sunset: 410 Gone, and the kling-3 id on Sume
Krea retired Kling 1.0 through 2.1: endpoints return 410 Gone after Sep 15, 2026. Sume lists kling-3 in its video catalog; here is what to change.

Kling models in Runway MCP vs Sume: what maps to what
Runway MCP added four Kling entries on Sep 18, 2026. Sume lists a Kling v3 Pro row and a Kling motion control MCP tool; it has no Kling O3 row.

Korea AI Basic Act Article 31: labeling an exported AI video
MSIT's guidelines ask for a human-recognizable label when AI video leaves the service. Here is how to burn a visible Korean label onto a clip with cues.

LangGraph interrupt response_schema for paid video approval
LangGraph 1.2.12 adds response_schema to interrupt(). Use a typed approve, edit or reject reply with max spend to drive a Sume dry_run, then submit.

LlamaIndex failed function tools: decide Sume retries in the tool
LlamaIndex v0.14.25 stops retrying failed function tools. Decide inside your Sume tool which errors are safe to retry, and return a status string.

Make webhook "Queue is full" 400: what Sume does and how to recover
When a Make webhook answers 400 Queue is full or 429, Sume retries up to 10 times at a fixed gap. If all fail, redeliver the job's event once the queue drains.

Mastodon alt text: 10,000 characters and a video transcript
Mastodon 4.6.0 raised the media alt text limit to 10,000 characters. Seed a long video description from a Sume transcript and a few stills.

Mastra background tasks and Sume: reuse the idempotency key
Mastra 1.72.0 fences background tasks with leases. If a new owner restarts a Sume tool call, the same idempotency key returns the original job.

MCP stream dropped: re-issue with the same idempotency_key
MCP 2026-07-28 says re-issue a broken request with a new request ID. On a paid Sume tool, keep the same idempotency_key so the retry is not a second charge.

MCP OAuth token expires: Sume's one-hour token, no refresh
Sume's hosted MCP OAuth access tokens last one hour and the server advertises only the authorization_code grant, so re-login or use an API key for long runs.