Guides
Sume guides: the API, Formats, agents, models, media tools and pricing, each answer dated and sourced.
MCP max_spend_exceeded: how Sume compares a dry run to your cap
max_spend_usd is optional on Sume's paid MCP tools, from 0 to 10,000. If the estimate is higher, max_spend_exceeded stops the call before it bills.
Meta political ads: disclose photoreal AI video and realistic audio
Meta's rules for social issue, election and political ads require disclosing photorealistic AI images, video and realistic audio. What counts in a Sume clip.
Microsoft's Content Provenance Detection: what you can check on a file
Foundry has a detection website and API for provenance. What the page says it checks, its limits, and how to use it on an AI clip or image from any generator.
Microsoft: C2PA may not survive a crop, transcode or compression
Microsoft's provenance page lists the edits that can drop credentials and watermarks. Which of them a Sume trim, filter or caption job could be, and a check.
Solo creator can't top up Sume? The team-wallet-only error
Sume top-ups are bought from a team wallet. If you see stripe_topup_team_only on a personal account, here is what the message means and the two ways out.
Audio transcript to timestamped Markdown notes: Python with Sume STT
Turn a 10-minute recording into Markdown with [mm:ss] sentence lines for Notion or Obsidian using Sume STT sentence segments. Cost is about 10 cents per file.
Turn STT words into paragraphs: break at pauses over 1.2 seconds
Sume STT returns a flat text string. Use word start and end times to break it into paragraphs at long pauses. Python, no extra API call or cost.
TikTok AI labels now say why: detection, creator or TikTok's tools
TikTok's Nov 2025 update adds context on whether a label came from AI detection, a creator label or a TikTok AI tool. Which one your Sume clip will show.
TikTok has labelled 1.3 billion videos: what that means for your ads
TikTok says it had labelled over 1.3 billion videos by Nov 2025 and uses C2PA Content Credentials. What an ad account should assume about AI clips.
A/B test two AI music beds on one ad: cost and the Timeline steps
Make two beds for one ad at $0.125 each, render both cuts on Timeline at $0.10 per minute, and compare. A cost table for a 30-second ad.
Afrobeats background music for a fashion ad: rhythm brief and cost
Brief an instrumental Afrobeats bed for a fashion ad: tempo, percussion, bass, and a short loopable bar. $0.125 per take on Sume, mixed on Timeline for $0.10.
AI Avatar Museum Docent Clip: Script, Cost and Limits
A museum exhibit docent clip as an AI avatar render: a 3-scene script, the seconds math, request JSON and cost on Sume, plus why it is not a live guide.
AI jingle from a prompt: a Lyria 3.5 use case on Sume with a trim
Google lists jingles and ringtones as Lyria 3.5 uses. On Sume, prompt a 10-second jingle for $0.125, then trim it with Timeline audio split for $0.01.
Animal Shelter Adoption Appeal as an AI Avatar Video: Cost
A shelter adoption appeal using an avatar and a scene photo: scripted scenes, per-scene seconds, request shape and Sume cost on each quality tier.
Anki cards with AI images: collection.media and the img tag
Anki finds images by filename in collection.media, via an img tag, with Allow HTML on. Make one PNG per word with Sume and write a tab-separated import file.
Astro Image with a Sume URL: allow the domain, set width and height
Astro only optimizes remote images from authorized sources and requires width and height on them. Add the host to image.domains, or save the Sume PNG under src.
Avatar Video Scene Durations: Fields Must Agree and the 2.8 Rule
How Sume sets a text scene's length: ceil(words / 2.8), the 0.001-second agreement rule across duration fields, and worked examples of 14 and 15 words.
Baby announcement video music: a soft lullaby-style bed on Sume
Brief a gentle instrumental for a baby announcement video: music box, soft piano, slow tempo. A take is $0.125 on Sume; fade it with Timeline.
Best API for text to speech: six checks to run before you pick
Compare TTS APIs on price per million characters, request cap, latency, languages, controls and word timing, with MAI-Voice-2.1 and Sume TTS 1.0 numbers.
Bulgarian, Croatian, Slovak text to speech API: one request each
Cartesia's Sonic 3.6 list includes bg, hr and sk. Run three scripts through Sume TTS 1.0 with a language map. Three 2,000-character jobs quote 30 cents.
Country and acoustic folk music for a brand ad: a warm brief
Brief an instrumental country-folk bed for a brand ad on Sume: guitars, pace, and a warm ending. A take is $0.125; a 30-second ad renders at $0.10.
Crowdfunding video music for a 2 to 4 minute pitch without fatigue
A 3-minute pitch video needs music that shifts with the story. Brief three Lyria 3.5 cues at $0.125 each, join with Timeline audio concat at $0.01, then duck.
C# HttpClient and Sume images: keep the key off the download call
Reuse one HttpClient, but set the bearer header per request, not as a default header, or it also rides along when you download the Sume image URL.
Do resizing or color correction need Meta's AI info label?
Meta says minor edits such as resizing or color correction do not need an AI label, and not every ad shows one yet. What that means for a generated ad clip.
Driving School Lesson Reminder as an AI Avatar Video: Cost
Remind learners about lessons with an avatar video: script, seconds per scene, request JSON and the Sume cost per clip on standard, plus and max.
ElevenLabs Music 'Commercial, no streaming': what an ad can use
ElevenLabs lists Starter as 'Commercial, no streaming' and Creator as 'Commercial, not enterprise'. What its terms exclude: film, TV, radio, and Studio Games.
Free video plans compared: Synthesia, HeyGen, Headliner, Sume
Synthesia Basic gives 500 credits, HeyGen Free 3 videos, Headliner Free 1 audiogram. What Sume's $0 plan includes and what it does not.
Gaming montage music at 140 BPM: an electronic brief with a drop
Brief a 140 BPM electronic track for a gaming montage on Sume. A $0.125 Lyria 3.5 take, a drop at a timestamp, and a Timeline render at $0.10 a minute.
GitHub README banner with light and dark versions from Sume
Generate two 3:1 banners with a 1536x512 custom size, commit them to the repo, and switch them with a picture element and prefers-color-scheme in the README.
Go: decode a Sume WebP image with golang.org/x/image/webp
Go's image package decodes PNG and JPEG out of the box; WebP needs golang.org/x/image/webp. Decode a Sume image URL and read its size without a re-encode.
Godot: import a Sume image as a texture, WebP or PNG
Godot imports PNG, JPEG and WebP. Drop a Sume image into res://, pick Lossless or VRAM Compressed, and use 1024x1024 where power-of-two sizes help.
Google Play: declare AI-generated store images and videos, per asset
Play Console asks you to declare AI-generated listing and promo images and videos, one asset at a time. How the checkbox works and what to log for Sume clips.
Google Sheets =IMAGE() with a Sume URL: PNG, JPEG, never SVG
=IMAGE needs a URL with a protocol and rejects SVG and drive.google.com links. Put a Sume data[0].url from a png or jpeg request in a cell and pick a mode.
Google Slides createImage with a Sume URL: use PNG or JPEG
CreateImageRequest takes PNG, JPEG or GIF, up to 50 MB and 25 megapixels, from a public URL of 2 KB or less. Ask Sume for output_format png or jpeg, not WebP.
Hardware Store DIY Tip Video With an AI Avatar: Cost per Clip
Weekly DIY tips from a hardware store host avatar: a short scene script, seconds per scene, request JSON, and Sume cost per clip on each quality tier.
Hip-hop beat for a product ad: instrumental brief and cost
Brief an instrumental hip-hop beat for a product ad with BPM, drums, bass, and sample feel. One take is $0.125 on Sume; mix it under a voiceover with duck_db.
Hugo: process a Sume image with Fill and a WebP output spec
Hugo reads PNG, JPEG and WebP and can write WebP itself. Save a Sume PNG into the page bundle, then use Fill to cut a 1200x630 card and output WebP.
HVAC Seasonal Maintenance Reminder as an AI Avatar Video
A seasonal HVAC tune-up reminder rendered as a Sume avatar video: script, seconds per scene, request JSON, and cost per clip and per customer list.
ImageMagick: turn a Sume WebP into a 1080x1350 JPEG
One magick command resizes to fill, center-crops to 1080x1350 and writes a stripped JPEG. Use it on a Sume 4:5 image whose native size is not exactly 1080x1350.
Is an unedited AI clip original enough for Shorts?
A raw AI clip posted as-is: what YouTube's pages say, what they skip, and a short list of cheap edits that add your own voice. Costs from Sume docs.
K-pop style instrumental for a product video: a brief with no artist
Brief Lyria 3.5 on Sume for a K-pop style instrumental by tempo, synths, and drops, not artist names. $0.125 per take; Timeline mixes it at $0.10 a minute.
Kickstarter Update Video With an AI Avatar and Product Image
Use an avatar to present a crowdfunding update with a product_image: script, seconds, request fields and Sume cost with and without the product image rate.
Make a Talking Photo Speak Spanish: TTS Language, Then Fabric
Two steps on Sume: generate Spanish speech with the language set, then animate a still with Fabric. Costs for a 30-second clip and what Avatar Video can't do.
MCP jobs_wait wait_deadline_exceeded vs wait_canceled: what to do
Both are retryable 503 results from Sume's jobs_wait with next_action poll_status. The job is untouched: call jobs_wait again, never resubmit.
MCP missing_tool_argument vs invalid_tool_argument: three look-alikes
Sume's MCP server has three near-identical codes: missing_tool_argument, invalid_tool_argument and invalid_tool_arguments. Learn which field each one names.
MCP 503 mcp_oauth_unavailable vs 401: do not re-sign-in
A bad OAuth token gets 401 with WWW-Authenticate; a server fault gets 503 mcp_oauth_unavailable or mcp_oauth_not_configured. Retry on 503, sign in only on 401.
Which Foundry models add C2PA and a watermark: GPT Image, FLUX.2, MAI
Microsoft's Foundry page lists the image, audio and voice models that get C2PA credentials and watermarks. What that does and does not tell you about Sume.
MiniMax H3 aigc_watermark defaults to false: what Sume exposes
MiniMax's H3 API has an aigc_watermark boolean, off by default. What it does, and what to do for a minimax-h3 clip made through Sume.