Guides
Sume guides: the API, Formats, agents, models, media tools and pricing, each answer dated and sourced.
Video to video motion transfer with AI: MiniMax H3 reference video
MiniMax H3 lists V2V motion transfer. Send a motion video and character images as references on Sume, write each reference's job, and mind the limits.
MiniMax H3 voice reference: match a voice with an audio clip
MiniMax H3 accepts up to 3 audio clips as references. How to write the prompt, the limits Sume enforces, and the rule that audio cannot be the only reference.
Mochi video model: what Mochi 1 is, needs and can't do
Mochi 1 preview is Genmo's Apache 2.0 open video model: 480p output, about 60 GB VRAM on one GPU, and it is not in Sume's video catalog.
fal subscribe equivalent for a video API: Sume has no SSE
Sume accepts mode subscribe but it is an alias of sync: one 30-second wait, no event stream. For a fal-style long wait, submit async and poll status.
Music API waveform data: get peaks from the audio file yourself
ElevenLabs added optional waveform data to music generation. Sume's music result has no waveform field, so compute peaks from the audio artifact with ffmpeg.
n8n Agents: give an agent a Sume video workflow
n8n's new Agents can use workflows and MCP servers as tools. A workflow that calls Sume keeps the API key out of the agent's hands.
n8n AI agent HTTP Request tool for video jobs
Attach the HTTP Request node to an n8n AI agent as a tool and use Optimize Response to hand the model only a job's status and result URL.
n8n AI Agent: force tool call on first iteration with Sume
n8n 2.40 added Force Tool Call on First Iteration to AI Agent v3. If Sume MCP is the tool, the first call is real: what a session can reach and the gates.
n8n MCP custom headers with OAuth token refresh: Sume caution
n8n 2.40 keeps custom headers on MCP registry remotes through OAuth token refreshes. With Sume, do not add an API-key header on top of an OAuth bearer token.
Nano Banana 2 Lite API: what Google shipped and what Sume lists
Nano Banana 2 Lite is Google's gemini-3.1-flash-lite-image model. What Google says it does, and which Nano Banana models Sume's image API lists today.
Nano Banana 2 pricing per image on Sume, by resolution tier
What one Nano Banana 2 image costs on Sume at 512, 1K, 2K and 4K, how the price scales with n, and where Google publishes its own per-image rates.
Nano Banana 3 API key and cost: what you can buy today
No Nano Banana 3 price exists on the Google pages read. What Google lists for Nano Banana 2 Lite, and how to read the cost of Sume's Nano Banana models.
Nano Banana consistent characters and 14 objects: the Sume cap
Google says Nano Banana 2 keeps up to 5 characters and 14 objects consistent. Sume takes 10 references per request, so 19 subjects need a plan.
Nano Banana 4K resolution API: the tier strings Sume accepts
Google says Nano Banana 2 spans 512px to 4K. On Sume you pick a tier with resolution: 512, 1K, 2K or 4K. The default, and why 4K can return 202.
Nano Banana character consistency: reference limits and a Sume request
Google documents character reference slots for its Nano Banana models. The limits, and how to keep a character across images with input_references on Sume.
Nano Banana Pro custom aspect ratio: what Sume does
Nano Banana Pro on Sume takes 11 native aspect ratios. A custom ratio or WxH size is snapped to the nearest one. Which, and how to check.
Nano Banana Pro vs Nano Banana 2: which id to send on Sume
Nano Banana Pro and Nano Banana 2 share tiers and 10 references on Sume; they differ in extra aspect ratios and in which tiers change the price.
Nano Banana quality parameter: why quality returns 400
Sending quality to Nano Banana 2 or Pro on Sume returns 400 unsupported_parameter. What quality does elsewhere, and what to change instead.
Nano Banana watermark: SynthID and C2PA on images via Sume
Google says Nano Banana 2 images carry SynthID and C2PA Content Credentials. Sume's docs say nothing about either, so check the file you download.
Netlify function timeout: run an AI video job without waiting
Netlify functions stop at 60 seconds, scheduled ones at 30, background ones at 15 minutes. Submit the video job and take the result on a callback.
Notion button send webhook: put the video URL back on the page
A Notion button can send a webhook that starts a Sume video job. When Sume calls back, PATCH the page's URL property with the finished video.
Open source AI video generator with audio: LTX-2 and MiniMax H3
Two open-weights video models make sound with the picture: LTX-2 and MiniMax H3. What each page says, and how a hosted API reports audio.
Open-source video model vs API: which to use for Wan-class video
Weights you run yourself, or a hosted video API? What the choice changes for cost, setup and limits, with Wan 3.0's per-second API price as the worked example.
OpenAI Agents API remote MCP server: what to give it for Sume
OpenAI's Agents API takes MCP servers at session setup. Give it Sume's mcp.sume.com/mcp URL and an API key, and gate paid calls with idempotency_key.
OpenAI Agents API webhooks: when the agent finishes, and Sume's
OpenAI says to stream output or use webhooks to learn when an agent finishes or needs input. Sume sends one agent.run.terminal webhook per run, never on cancel.
OpenAI Agents SDK human in the loop for paid tools
Set needs_approval on a function tool that starts a paid video job. The run pauses with result.interruptions until you approve or reject the call.
Restrict API key creation: what OpenAI added, and Sume's key rules
OpenAI added admin controls to restrict API key creation on 2026-09-15. Sume keys are workspace-scoped with fixed scopes and rotate by replacement.
Pinterest video ad specs: standard and Max Width ratios
Pinterest's product specs: MP4, MOV or M4V up to 2 GB, 4 s to 15 min, standard ratios 1:2 to 1.91:1, Max Width no taller than 1:1. Sume settings.
Audio podcast to video with AI: cover art plus the audio
Turn podcast audio into a video by holding one cover image for the whole episode on Timeline 1.0, then cut short clips with an audio split.
PowerPoint to video AI: turn a slide deck into a narrated video
To turn a PowerPoint into a video with AI, export slides as images and speaker notes as text. Sume's agent takes up to 30 image attachments and notes in input.
Premiere Pro AI video plugin: what Sume offers instead
Sume's docs describe no Premiere Pro plugin. Its route is an API loop that submits shots, waits for each job and downloads MP4 files for a Premiere bin.
AI avatar video preview before render: what you pay when
Sume's docs publish no price for an avatar video preview. The render is billed per second by tier, and the tier is chosen at generate-video, after you approve.
Qwen Image 2.1 API: open weights, and what Sume lists
Qwen-Image-2.1 is an open-source 7B image model on Hugging Face with no inference provider. Sume lists qwen-image and qwen-image-max instead.
Qwen Image 3.0: does Sume list it? A catalog check
Sume lists qwen/qwen-image and qwen/qwen-image-max, and nothing named 3.0. How to check the live catalog for a new Qwen id and what an unknown id returns.
Qwen Image Edit 2511: what Sume offers for image edits
Sume lists no model named Qwen Image Edit 2511. It does take edits on qwen/qwen-image with up to 10 reference images. The request, the limits, and the gaps.
Qwen Image Layered: what Sume returns, flat or transparent
Sume's image API returns flat images, not separate layers. It lists qwen/qwen-image and qwen/qwen-image-max, and offers transparent backgrounds on other routes.
Qwen Image license: is Qwen-Image open source under Apache 2.0?
The Qwen-Image model card says the weights are Apache 2.0 and were released on 2025-08-04. What that covers, and how the hosted qwen ids on Sume differ.
Qwen Image Max API on Sume: text-only, and how it differs
qwen/qwen-image-max on Sume is text-to-image only and rejects reference images with a 400. Price, ratios, a working call, and how it differs from qwen-image.
Recraft V4.1 API on Sume: check the catalog for the id
Sume lists recraft/recraft-v4, not a V4.1 id. How to check the live catalog for a newer Recraft, and what the listed model takes: text only, webp, $0.05.
Recraft V4 SVG: can you get vector output through Sume?
Recraft says V4 can make editable SVG. Sume's recraft/recraft-v4 advertises webp output only. What that means, plus a working call.
Reddit ad image size: 1080x1080, 3 MB, JPG or PNG
Reddit image ads take JPG, PNG or GIF up to 3 MB. It recommends 1080x1080 (1:1), 1080x1440 (3:4) and 1080x1350 (4:5). How to ask Sume for those ratios.
Reddit video ad specs: MP4 or MOV, 1 GB, 30 FPS
Reddit video ads take MP4 or MOV up to 1 GB at 30 FPS or less, 2 seconds to 15 minutes, with sizes like 1200x1200 and 900x1200. How to render to them.
Instagram Reels caption limit: 2,200 characters, 30 hashtags
Meta's Graph API lists 2,200 characters, 30 hashtags and 20 @ tags for a Reel caption. This is different from burned-in video captions, which Sume adds.
Reels cover photo size: 9:16 JPEG, 8 MB, and how to make one
Meta's Instagram Graph API says a Reels cover is a JPEG up to 8 MB, 9:16 recommended, and cover_url beats thumb_offset. How to make one with Sume.
Extract on-screen text from a reference video with an API
Reference ingest reads on-screen text at source resolution, returns lines with boxes, spans and confidence, and flags low-confidence lines instead of guessing.
Check whether a reference video is silent before adding music
Reference ingest reports audio.silent at a -60 LUFS gate, speech presence and beats, so you know whether to keep, replace or add a soundtrack before a remix.
Remove the background and upscale an image: which order?
Upscale first, then remove the background, so the file you keep is the RMBG PNG with alpha. Sume's docs do not say the upscaler keeps transparency.
remove.bg API is moving to Leonardo: a background removal API
remove.bg says its API moves to Leonardo.Ai on December 1, 2026. If you need a background removal API, here is Sume's RMBG 1.0 contract.