xW 3.0 S Pro

REQUIRED. Email of YOUR real end-user inside your product — the person who actually triggered this video. We use it to attribute every request and give you a chain of responsibility: you can list, audit, and delete that user's videos. Do NOT pass a fake address, a shared placeholder, or someone else's email — that's a policy violation.

Example: "alice@your-product.com"

Public model alias. Required — pass the alias of the video model you want.

Example: "xW3.0sPro"

What the video should contain and how it should move. Optional on this model: a request that supplies a keyframe or reference media may omit it. Refer to reference assets positionally — "Image 1", "Video 2", "Audio 1" — numbering each kind separately in the order you listed them.

Example: "Image 1 walks through the doorway in Image 2, camera slowly pushing in"

Image pinned as the exact FIRST frame of the output. Keyframe mode — cannot be combined with any `reference_*` field.

Example: "https://example.com/opening.png"

Image pinned as the exact LAST frame of the output. Pair it with `first_frame` to have the model invent the transition between two fixed states. Keyframe mode.

Example: "https://example.com/closing.png"

Reference images the model draws on. Numbered "Image 1", "Image 2" … in array order for use in the prompt. Reference mode — cannot be combined with `first_frame`/`last_frame`.

Example: ["https://example.com/character.png"]

Reference video clips, numbered "Video 1", "Video 2" … in array order. Reference mode. Billed: the combined length of the clips is added to the output length, and the total is charged at this request’s per-second rate.

Example: ["https://example.com/motion.mp4"]

Reference audio clips, numbered "Audio 1", "Audio 2" … in array order. Reference mode.

Example: ["https://example.com/voice.mp3"]

Output video resolution. Determines the per-second rate.

Example: "1080p"

Aspect ratio of the output. `adaptive` lets the model pick one that suits the input material and the prompt — usually the right choice when you pass frames or references of a known shape.

Example: "16:9"

Seconds of output video, billed per second. When the request carries reference video, this plus the combined reference length may not exceed the combined ceiling — see Limits.

Example: 10

Whether the output carries an audio track. The model generates dialogue, effects and ambience described in the prompt. Turning it off does not change what you are charged.

Example: false

Run on the faster tier. The model and the output quality are identical — only latency differs, and substantially: a long clip finishes in roughly a quarter of the time. It costs more per second, so it is a deliberate trade rather than a default. Both rates are published by `client.models.list()`.

Example: true

Determinism seed (0–2147483647). Pass null or omit for an auto-generated seed. Same seed with the same prompt, model and parameters reproduces the same video.

Example: 42

Optional public HTTPS URL we will POST the final record to once the video is succeeded or failed. The body is identical to what GET /v1/videos/:requestId returns. Each delivery is HMAC-signed (`X-XMode-Signature: v1=<hex>`) and timestamped (`X-XMode-Timestamp`); reject anything older than 5 minutes. Private/loopback IPs are refused (SSRF guard).

Example: "https://your-app.example.com/hooks/xmode"

Unique id of the video request. Use it with GET /v1/videos/:requestId to poll.

Public alias of the model that will produce the video.

Lifecycle marker. POST always returns `queued`. GET returns the current state. `expired` means the request succeeded more than 23 hours ago and signed URLs no longer work — re-run if you need the video again.

Present on `queued` responses only. Absolute URL to GET for status. Recommended polling cadence: every 5 seconds until terminal.

Present only when status=`succeeded`. Each item has `id`, `expiresAt`, optional `duration`/`resolution`, and EITHER `url` (signed, valid 23 h) OR `error: { code: "storage_failed" }` if our storage layer dropped it. If every video of the request is dropped there is nothing to deliver, so the request comes back `failed` and refunded instead. The URL is re-signed on every read, so re-reading this record replaces a link that went stale on your side — but that never moves `expiresAt`, which is fixed when the generation completed. Once it passes, the record returns status `expired` with no URL and the bytes are gone. Download inside the window or re-generate.

Present only when status=`failed`. `{ code, message }`. Same code values as the top-level error format (validation_error, content_policy, provider_error, provider_timeout, internal_error).

Present only when status=`succeeded`. `{ xTokens: number }`. Debited at request time; refunded automatically on `failed`.

Present on terminal statuses (`succeeded`, `failed`, `expired`).

Echoed back so you can confirm the grouping. Always lowercase.

When the request was accepted (POST time).