Agencies and studios can’t afford to guess what resonates. Cue measures it.
A creative brief goes in. A grounded answer comes out — the cue, the cast, the market fit, the price and the path to signature — every ranking traced to measured audience affinity in the Qloo Taste Graph. Then Cue does the thing a ranking cannot do for itself: it measures how much of the brief’s cultural taste space the slate actually reaches, and rebalances it when eight defensible picks still serve one corner of the audience.
- Engine
- Qloo Taste Graph
- Endpoints per run
- —
- Mode
- ready
Scene brief
Paste the script or a plain description. Named films and artists become Qloo signal entities.
Loading the agent…
Cue will run these through /v2/analysis/compare to see whether the two readings actually split the audience.
How a cue sheet gets made
The model reads the scene. Qloo decides what ranks.
- 01you
Your scene
A script, treatment or plain description. Named films and artists become anchors.
- 02model
Scene reader
Reads tone, era and instrumentation into a structured brief. Invents no numbers.
- 03Qloo
Qloo Taste Graph
Resolves the anchors, then scores every candidate by measured audience affinity.
- 04Qloo + model
Evidence-backed sheet
Ranked cues, audience fit and price — each carrying the Qloo signal behind it.
Nothing scored yet
Pick a sample scene or paste your own, then build the cue sheet. Cue will call the Qloo Insights API, rank artists by measured affinity, check them against the target audience, and show you every call it made.
Cross-domain ranking
Film taste predicts music taste. Qloo traverses it; an LLM cannot.
Explainable picks
Every cue names the signal that caused it.
Audience-checked
Demographic signals, not vibes.
Full receipts
Endpoint, parameters, latency, result count.