What every billable feature charges your workspace, in both units (1
credit is 1000 tokens) and per what: a minute of video, an image, or a
single generation. Use it to estimate a job before queuing it. A
workspace created from 2026-10-26 pays one Video charge per minute of
video in place of the render, captions, scenes and premium voice
charges, so its list differs.
Authorization
bearerAuth An API key created under Settings > API & Webhooks. Keys start with sf_live_.
In: header
Response Body
application/json
application/json
curl -X GET "https://example.com/reference/credit-costs"{ "data": [ { "label": "Video render", "tokens": 1800, "credits": 1.8, "per": "minute", "note": null }, { "label": "Subtitles & B-rolls", "tokens": 1200, "credits": 1.2, "per": "minute", "note": null }, { "label": "AI image", "tokens": 1000, "credits": 1, "per": "image", "note": null }, { "label": "Script or description", "tokens": 80, "credits": 0.08, "per": "generation", "note": null }, { "label": "Premium voice", "tokens": 3000, "credits": 3, "per": "minute", "note": "on top of the standard voice" } ]}List aspect ratios GET
The values `brief.aspectRatio` accepts, with what each one is usually for.
List job failure codes GET
The `error.code` values a failed job reports, with a human title, the `action` that clears each one (`upgrade`, `credits`, `connect`, `configure`, `schedule`, `retry`, `trial` or `none`) and `blocked`: an expected limit (`warning-*`, the same codes a `400` carries when the limit is hit synchronously) rather than a failure.