Reads the feed
Cue ingests briefs from RSS, webhooks, CMS APIs, product catalogues, and scheduled queries. Every item lands as a normalised brief ready for production.
Cue is the intelligence inside the engine. It drafts the script, plans the shots, reads your notes like a human editor would, and shows you what it wants to change before it spends a credit.
/source review feed priorities
Your CMS sent 14 articles today. Three are tagged breaking, but your audience clicks the explainers 2.4x more on Reels. Want me to flip the priority for the rest of the week?
/template pause underperformers
Your News Brief template has been winning. Your Listing Tour template hasn't hit baseline in 11 days. Want me to pause it?
Cue ingests briefs from RSS, webhooks, CMS APIs, product catalogues, and scheduled queries. Every item lands as a normalised brief ready for production.
News brief, product video, listing tour, recap reel, changelog explainer. Cue picks the right schema, length, voice, and visual treatment for the channel the brief is bound to.
Storyboard, script, voice, music, imagery, edit, render. Cue owns every stage and surfaces only the decisions that need a human in the loop.
Per-channel voice, palette, watermark, music mood, lower-third style. Cue enforces brand defaults on every output without asking twice.
One master render, automatic 9:16, 16:9, and 1:1 variants. YouTube, TikTok, Instagram Reels, LinkedIn, X. Cue handles platform metadata and posting cadence per channel.
Cue ingests platform analytics, scores templates against channel baselines, runs thumbnail and hook A/Bs, and recommends source weightings the next time the feed updates.
When a template stops paying for itself or a source feeds noise, Cue says so, proposes the next move, and pauses the bleed on channels where you have authorised her to.
An assistant that spends your money without asking is not an assistant. These are the rules Cue runs under, on every video.
Words, voice, look, scenes, cut. Cue stops at each one and waits for you.
Down to the cent, on every regenerate. No surprise invoices.
Cue suggests rewrites as diffs. It never silently edits a line you approved.
Every accepted change is a version. Roll back to any of them for free.
No. Your briefs, source feeds, brand voice, rendered video, and platform analytics stay private to your workspace. Cue only reads your data during an active request, and nothing is shared, sold, or used for model training.
Both. The same Cue logic that powers ask-and-answer inside the app runs against the public API. POST a brief, receive a finished video plus a Cue recommendation for the next brief in the same callback payload.
Cue scores every template, source, and cadence against the channel baseline you configured. Recommendations cite the actual metric behind them and never fire on a single noisy data point.
Per channel you choose. By default Cue proposes and you approve. Authorise her on a channel and she will pause losing templates, rotate variants, and shift cadence inside the limits you set.
Publishers, newsrooms, e-commerce platforms, real estate platforms, podcast networks, agencies, and in-house brand teams. Anyone with an existing content stream who needs video as a missing modality at low marginal cost.
The free tier is a developer wedge with a small monthly video credit pool. Paid plans are per-video credit plus volume tier, with template marketplace revenue share for community-contributed templates. Slash-command runs inside the dashboard do not count against your video credits.
Cue replies in English today. Output video voiceovers cover 20+ languages, and full multilingual Cue conversations are a 2026 roadmap item.
Yes. Workspace settings let you disable proactive recommendations and keep Cue as a passive operator that only runs the brief you send. The engine still works, you just lose the optimise loop.