REST API Reference¶

Base URL: https://api-v5.anchor.fm

The anchor.fm internal REST API handles episode creation, metadata read/write, show settings, and media upload. It predates and coexists with the GraphQL API — Anchor's acquisition by Spotify left this backend intact.

Common Headers¶

All requests require:

Authorization: Bearer {bearer_token}
Accept: application/json

POST / PUT requests additionally require:

Content-Type: application/json
Origin: https://creators.spotify.com
Referer: https://creators.spotify.com/

Common Query Parameters¶

Parameter	Description
`isMumsCompatible=true`	Required on every endpoint. Never omit.
`returnWebIds=true`	Recommended on `/overview`. Omitting it may drop some fields.

ID System¶

The anchor.fm API uses anchor numeric IDs (integers), not the Spotify string IDs visible in the S4C UI. You need to convert between them before calling most endpoints.

ID type	Example	Used in
Spotify Episode ID	`"1xouj0WrH2klavXKzDWZbq"`	S4C UI, GraphQL
Anchor Episode ID	`123456789`	anchor.fm REST endpoints
Spotify Show ID	`"YOUR_SHOW_ID"`	S4C UI, GraphQL
Station ID (numeric)	`YOUR_STATION_ID`	anchor.fm REST endpoints

1-A. Spotify Episode ID → Anchor Numeric ID¶

GET /v3/episodes/spotify:episode:{SPOTIFY_EP_ID}/episodeId?isMumsCompatible=true

Translates a Spotify Episode ID into the numeric ID required by all other episode endpoints.

Response¶

{ "episodeId": 123456789 }

PythoncURL

resp = requests.get(
    f"https://api-v5.anchor.fm/v3/episodes/spotify:episode:{spotify_ep_id}/episodeId"
    "?isMumsCompatible=true",
    headers=headers_get,
)
anchor_id = resp.json()["episodeId"]

curl -s \
  -H "Authorization: Bearer $BEARER" \
  "https://api-v5.anchor.fm/v3/episodes/spotify:episode:${SPOTIFY_EP_ID}/episodeId?isMumsCompatible=true"

1-B. Get Episode Overview¶

GET /v3/episodes/{ANCHOR_ID}/overview?isMumsCompatible=true&returnWebIds=true

Returns the current metadata for an episode.

Response fields¶

Field	Type	Description
`title`	string	Episode title
`description`	string	Description HTML
`isPublished`	boolean	Whether the episode is published
`publishOn`	string \| null	Scheduled publish time (UTC ISO 8601). Present only for scheduled episodes
`userId`	integer	Account user ID — required for update calls
`podcastEpisodeType`	string	`"full"` / `"trailer"` / `"bonus"`
`podcastEpisodeIsExplicit`	boolean	Explicit content flag

Episode state matrix¶

`isPublished`	`publishOn`	State
`false`	present	Scheduled
`true`	—	Published
`false`	absent	Draft

Draft episodes return 403

Calling this endpoint on a draft episode (isPublished: false, no publishOn) returns HTTP 403 Forbidden — even in a live browser session. The S4C UI fetches draft data via GraphQL instead. If you need to update a draft episode's metadata, keep the original values in memory before calling the update endpoint.

PythoncURL

r = requests.get(
    f"https://api-v5.anchor.fm/v3/episodes/{anchor_id}/overview"
    "?isMumsCompatible=true&returnWebIds=true",
    headers=headers_get,
)
overview = r.json()

curl -s \
  -H "Authorization: Bearer $BEARER" \
  "https://api-v5.anchor.fm/v3/episodes/${ANCHOR_ID}/overview?isMumsCompatible=true&returnWebIds=true"

1-C. Update Episode (Title / Description / Publish Date)¶

POST /v3/episodes/{ANCHOR_ID}/update?isMumsCompatible=true

Updates episode metadata. You can change the title, description, episode type, explicit flag, and scheduled publish time in a single call.

Request body¶

{
  "userId": YOUR_USER_ID,
  "title": "Episode title",
  "description": "<p>Episode description HTML</p>",
  "episodeType": "full",
  "isPublished": false,
  "podcastEpisodeIsExplicit": false,
  "publishOn": 1748732400,
  "wizardDraftedToPublishOn": "2026-06-01T21:00:00.000Z"
}

Parameter	Type	Required	Description
`userId`	integer	Yes	Copy directly from the `/overview` response
`title`	string	Yes	Omitting it risks the server setting a blank title
`description`	string	Yes	HTML format
`episodeType`	string	Yes	`"full"` / `"trailer"` / `"bonus"`
`isPublished`	boolean	Yes	Pass the current value from `/overview` to preserve state
`podcastEpisodeIsExplicit`	boolean	Yes	Same as above
`publishOn`	integer	No	Unix timestamp (seconds). Required when scheduling. See warning below
`wizardDraftedToPublishOn`	string	No	UTC ISO 8601. Must match `publishOn` in wall-clock time

Response¶

HTTP 200 on success:

{
  "description": "<p>Episode description HTML</p>",
  "title": "Episode title",
  "didChangePublishState": true,
  "shareUrl": "https://creators.spotify.com/pod/show/..."
}

Scheduling behaviour

Sending isPublished: false together with a publishOn value causes the server to automatically treat it as isPublished: true (scheduled for future publication). This is the correct way to schedule an episode.

Published episodes lock publishOn

Once an episode is published (isPublished: true), you cannot change publishOn via this endpoint. Use the S4C UI or Playwright to modify the publish date of a live episode.

publishOn must be a Unix timestamp integer

publishOn expects a Unix timestamp in integer seconds, not an ISO 8601 string. Sending an ISO string (e.g. "2026-06-01T21:00:00.000Z") causes the server to parse only the leading digits as an integer — "2026" becomes 2026 seconds, storing 1970-01-01T00:33:46.000Z.

wizardDraftedToPublishOn is the opposite: it requires an ISO 8601 string. Sending an integer returns HTTP 400.

from datetime import datetime, timezone, timedelta
JST = timezone(timedelta(hours=9))
publish_jst = datetime(2026, 6, 1, 6, 0, 0, tzinfo=JST)
pub_utc = publish_jst.astimezone(timezone.utc)
publish_on_unix = int(pub_utc.timestamp())                    # → 1748732400
publish_on_iso  = pub_utc.strftime("%Y-%m-%dT%H:%M:%S.000Z") # → "2026-05-31T21:00:00.000Z"

PythoncURL

pub_utc = publish_jst.astimezone(timezone.utc)
payload = {
    "userId": overview["userId"],
    "title": "New episode title",
    "description": "<p>New description</p>",
    "episodeType": overview.get("podcastEpisodeType", "full"),
    "isPublished": overview.get("isPublished", False),
    "podcastEpisodeIsExplicit": overview.get("podcastEpisodeIsExplicit", False),
    "publishOn": int(pub_utc.timestamp()),                    # Unix integer
    "wizardDraftedToPublishOn": pub_utc.strftime("%Y-%m-%dT%H:%M:%S.000Z"),  # ISO string
}
r = requests.post(
    f"https://api-v5.anchor.fm/v3/episodes/{anchor_id}/update?isMumsCompatible=true",
    json=payload,
    headers=headers_post,
)

curl -s -X POST \
  -H "Authorization: Bearer $BEARER" \
  -H "Content-Type: application/json" \
  -H "Origin: https://creators.spotify.com" \
  -H "Referer: https://creators.spotify.com/" \
  -d '{"userId":YOUR_USER_ID,"title":"New title","description":"<p>Desc</p>","episodeType":"full","isPublished":false,"podcastEpisodeIsExplicit":false,"publishOn":1748732400,"wizardDraftedToPublishOn":"2026-06-01T21:00:00.000Z"}' \
  "https://api-v5.anchor.fm/v3/episodes/${ANCHOR_ID}/update?isMumsCompatible=true"

1-D. Get Region¶

GET /v3/region?isMumsCompatible=true

Returns the user account's region. Called automatically when loading the episode detail page.

Example response: { "region": "JP" }

1-E. Get Sponsored Content Status¶

GET /v3/episodes/{ANCHOR_ID}/sponsoredContentStatus?isMumsCompatible=true

Returns whether the episode contains sponsored content.

Response: { "containsSponsoredContent": false }

1-F. Get Cover Art Colors¶

POST /MultiGetColor?isMumsCompatible=true

Returns the color palette extracted from an image URL. Called when rendering cover art in the UI.

Request body¶

{
  "color_queries": [
    { "url": "https://anchor-generated-image-bank.s3.amazonaws.com/...jpg" }
  ]
}

1-G. Get Episode Form Data (Batch)¶

GET /v3/episodes/pcdFormData?episodeUri[]=spotify:episode:{SPOTIFY_EP_ID}&isMumsCompatible=true

You can supply multiple episodeUri[] parameters to batch-fetch form data. Called when the episode list page loads.

1-H. Get Show Metadata¶

GET /v3/stations/{STATION_ID}/metadata?isMumsCompatible=true

Returns show-level information: title, description, categories, cover art, and distribution flags. Called when the show settings page loads.

The STATION_ID is the anchor numeric station ID (distinct from the user ID). You can obtain it via the ID conversion endpoint (1-L).

Response fields¶

Field	Type	Description
`rssFeedUrl`	string	RSS feed URL
`podcastName`	string	Show title
`podcastDescription`	string	Show description
`podcastAuthorName`	string	Author name
`podcastAuthorEmail`	string	Author email
`podcastImage`	string	Cover art URL (original size)
`podcastImage400`	string	Cover art URL (400 px)
`podcastCategory`	string	Category in `"Arts\\|Design"` format
`podcastLanguage`	string	Language code (e.g. `"ja"`)
`language`	string	Same as `podcastLanguage`
`isDistributedOnSpotify`	boolean	Whether the show is on Spotify
`isRssFeedDisabled`	boolean	Whether the RSS feed is disabled
`hasPublishedEpisode`	boolean	Whether at least one episode is published
`distributionEligibilityForPaywalls`	string	Paywall eligibility (`"eligible"` etc.)
`isStationOver30DaysOld`	boolean	Whether the show is older than 30 days
`isStationMarkedAsSpam`	boolean	Whether the show is flagged as spam

1-I. Get Partner IDs¶

POST /v3/{USER_ID}/partnerIds?isMumsCompatible=true

Returns partner IDs related to monetisation (ads, Fan Support, etc.). Called when the show settings page loads.

1-J. Get User Verification Status¶

GET /v3/users/verification?isMumsCompatible=true

Returns the account's identity verification state.

1-K. Onboarding Status¶

GET /v3/onboarding/redirectStatus/{USER_ID}/?isMumsCompatible=true
GET /v3/onboarding/steps/{USER_ID}/?userId={USER_ID}&isMumsCompatible=true

Returns onboarding progress for new users. Has no practical effect on established accounts.

1-L. Spotify Show ID → Anchor Numeric IDs¶

GET /v3/shows/{SPOTIFY_SHOW_ID}/legacyIds?isMumsCompatible=true

Converts a Spotify Show ID into the anchor.fm internal numeric IDs needed by most other endpoints.

Response¶

{
  "userId": YOUR_STATION_ID,
  "stationId": YOUR_STATION_ID,
  "webStationId": "112fef78c"
}

Field	Description
`stationId`	The numeric ID used in episode list, metadata, and other endpoints
`userId`	The numeric user account ID
`webStationId`	The URL slug (vanity ID)

PythoncURL

r = requests.get(
    f"https://api-v5.anchor.fm/v3/shows/{spotify_show_id}/legacyIds?isMumsCompatible=true",
    headers=headers_get,
)
ids = r.json()
station_id = ids["stationId"]
user_id    = ids["userId"]

curl -s \
  -H "Authorization: Bearer $BEARER" \
  "https://api-v5.anchor.fm/v3/shows/${SPOTIFY_SHOW_ID}/legacyIds?isMumsCompatible=true"

1-M. Update Episode List Index¶

POST /v3/stations/{STATION_ID}/episode-list-index?isMumsCompatible=true

Synchronises the episode list index. Called automatically by the browser when the episode list page loads.

Response: { "success": true }

1-N. Get Show Form Data¶

GET /v3/stations/{STATION_ID}/pcdForm?isMumsCompatible=true

Response: { "formUrl": "" }

1-O. Get Distribution Settings¶

GET /v3/stations/{STATION_ID}/distribution?isMumsCompatible=true

Returns the show's distribution and RSS status. Called when the episode list page loads.

Response fields¶

Field	Type	Description
`canEnableRss`	boolean	Whether RSS distribution can be enabled
`externalUrls`	object	E.g. `{"spotify": "https://open.spotify.com/show/..."}`
`spotifyDistributionStatus`	string	Spotify distribution status
`spotifyShowUri`	string	Spotify Show URI
`platforms`	array	List of active distribution platforms
`vanitySlug`	string	URL slug
`isEnterprisePodcast`	boolean	Enterprise plan flag

1-P. Get User Settings¶

GET /v3/users/{USER_ID}/settings/termsAndContactability?isMumsCompatible=true

Example response¶

{
  "termsOfServiceVersion": 1,
  "agreedToMinimumRequiredTermsOfService": true,
  "contactabilityStatus": "optedOut",
  "countryRule": "needsResponseOptInSelected"
}

1-Q. Update Sponsored Content Status¶

PUT /v3/episodes/{ANCHOR_ID}/sponsoredContentStatus?isMumsCompatible=true

Sets whether the episode contains sponsored content. Called automatically when saving an episode.

Request body¶

{
  "containsSponsoredContent": false,
  "publishOn": 1748732022
}

Parameter	Type	Required	Description
`containsSponsoredContent`	boolean	Yes	Whether the episode includes sponsored content
`publishOn`	integer	Yes	Publish time as a Unix timestamp in seconds (not ISO 8601)

Warning

publishOn here is a Unix timestamp (integer seconds), not an ISO 8601 string. Passing an ISO 8601 string will silently corrupt the value.

1-R. Sync Monetisation Lifecycle¶

POST /v5/monetization-lifecycle/{STATION_ID}/sync-and-fetch?isMumsCompatible=true

Synchronises and fetches the monetisation state (e.g. Spotify Partner Program). Called when the episode list page loads.

Key response fields¶

Field	Description
`products.SPOTIFY_PARTNER_PROGRAM.lifecycleState`	Partner program state
`products.SPOTIFY_PARTNER_PROGRAM.pendingRequirements`	Outstanding requirements

1-S. Create a New Episode¶

POST /v3/stations/{STATION_ID}/episodes?isMumsCompatible=true

Creates an empty episode record. This is the first step in the episode wizard — it allocates the episode ID before audio or metadata are set.

Request body¶

{ "hourOffset": -9 }

Parameter	Type	Description
`hourOffset`	integer	User timezone offset (JST = `-9`)

Response¶

{
  "episodeId": 120529038,
  "stationId": YOUR_STATION_ID
}

Three-step episode creation

POST /v3/stations/{STATION_ID}/episodes — allocate the episode record (this endpoint)
Upload audio/video via the media upload flow (6-A through 6-D)
POST /v3/episodes/{ANCHOR_ID}/update (1-C) — set title, description, and publish settings

1-T. Update Show Metadata¶

POST /v3/stations/{STATION_ID}/metadata?isMumsCompatible=true

Updates show-level metadata: title, description, author, category, language, and explicit flag. Called when the podcast settings page is saved.

Request body¶

{
  "podcastName": "Your Podcast Title",
  "podcastDescription": "Show description",
  "podcastAuthorName": "Author Name",
  "podcastCategory": "Arts|Design",
  "podcastLanguage": "ja",
  "podcastEpisodeIsExplicit": false
}

Parameter	Type	Required	Description
`podcastName`	string	Yes	Show title (max 100 characters)
`podcastDescription`	string	No	Show description (max 600 characters)
`podcastAuthorName`	string	No	Author name (max 80 characters)
`podcastCategory`	string	No	Category in `"Arts\\|Design"` format; fetch valid values from `/v3/settings/podcast-category/options`
`podcastLanguage`	string	No	Language code (e.g. `"ja"`)
`podcastEpisodeIsExplicit`	boolean	No	Explicit content flag

HTTP 200 returns the updated metadata object.

1-U. Update Vanity Slug¶

POST /v3/settings?isMumsCompatible=true

Updates the show's custom URL slug.

Request body¶

{ "webStationId": "your-vanity-slug" }

Warning

webStationId must be globally unique. Using an already-taken slug may return 409 Conflict.

1-V. Image Upload Flow (Cover Art / Episode Thumbnail)¶

Cover art and episode thumbnails share the same upload pipeline but use different endpoints. The pipeline is separate from the media upload pipeline (6-A through 6-D).

1-V-1. Get a Signed Upload URL¶

Show cover art:

GET /v3/stations/{STATION_ID}/images/upload/signed_url?isMumsCompatible=true

Episode thumbnail:

GET /v3/episodes/{ANCHOR_ID}/images/upload/signed_url?isMumsCompatible=true

Response:

{
  "url": "https://storage.googleapis.com/anchor-image-storage-production/{UUID}/{FILENAME}?X-Goog-Algorithm=...",
  "uploadId": "a46e21a4-1d2d-1686-1355-905e191a3bef"
}

1-V-2. PUT the Image to Google Cloud Storage¶

PUT {signed_url}
Content-Type: image/png   # or image/jpeg

No auth headers needed — the signature is embedded in the URL.

1-V-3. Notify Upload Completion¶

Show cover art:

POST /v3/stations/{STATION_ID}/images/upload/{UPLOAD_ID}/process_upload?isMumsCompatible=true

Body: {} or empty.

Episode thumbnail:

POST /v3/episodes/{ANCHOR_ID}/images/upload/{UPLOAD_ID}/process_upload?isMumsCompatible=true

Body:

{
  "episodeId": 120520683,
  "uploadType": "episode_image"
}

1-V-4. Poll for Completion¶

Show cover art:

GET /v3/stations/{STATION_ID}/images/upload/{UPLOAD_ID}/?isMumsCompatible=true

Episode thumbnail:

GET /v3/episodes/{ANCHOR_ID}/images/upload/{UPLOAD_ID}/?isMumsCompatible=true

Poll until status equals "completed" (typically a few seconds).

1-W. Monetisation Endpoints¶

Called from the monetisation page (/pod/show/{SHOW_ID}/monetize/...):

GET /v4/paywalls/{STATION_ID}/lifecycle-state?isMumsCompatible=true
GET /v6/wallet/{STATION_ID}/availability?isMumsCompatible=true
GET /v6/wallet/{STATION_ID}/?isMumsCompatible=true
GET /v3/paywalls/{STATION_ID}/?isMumsCompatible=true

These endpoints return paywall and wallet state. Used together with 1-R.

1-X. Show Settings Page Endpoints¶

Called from the podcast settings page (/pod/show/{SHOW_ID}/podcast/edit):

GET /v3/stations/{STATION_ID}/episodePage?isMumsCompatible=true
GET /v3/station/{STATION_ID}/episodes/featured?isMumsCompatible=true
GET /v3/stations/{STATION_ID}/hostRecommendations?isMumsCompatible=true

These return "Start Here" featured episode and host recommendation settings.

1-Y. Distribution Page Endpoints¶

Called from the distribution page (/pod/show/{SHOW_ID}/podcast/distribution):

GET /v3/onboarding/redirectStatus/{STATION_ID}/?isMumsCompatible=true
GET /v3/onboarding/steps/{STATION_ID}/?userId={USER_ID}&isMumsCompatible=true
GET /v3/stations/status?isMumsCompatible=true

/v3/stations/status returns a list of all shows associated with the currently authenticated user.

1-Z. Delete Episode¶

POST /v3/episodes/{ANCHOR_ID}/delete?isMumsCompatible=true

Permanently deletes an episode. Works on drafts, scheduled (not yet published) episodes, and published episodes alike.

Request body¶

{}

An empty JSON body is sufficient. The episode is identified entirely by the {ANCHOR_ID} path segment (the numeric Anchor ID, not the Spotify episode URI — resolve it first via 1-A if you only have the Spotify ID).

Response¶

{}

Returns 200 on success.

Irreversible

There is no trash/restore step — a successful call removes the episode immediately. Confirm the {ANCHOR_ID} resolves to the intended episode (e.g. via /overview, 1-B) before calling.

curl -X POST \
  "https://api-v5.anchor.fm/v3/episodes/{ANCHOR_ID}/delete?isMumsCompatible=true" \
  -H "Authorization: Bearer ${BEARER}" \
  -H "Content-Type: application/json" \
  -H "anchor-client-type: web" \
  -H "anchor-api-version: 3.8.3" \
  -d '{}'

Known Quirks¶

description field — server-side transformations¶

When the server stores the description field, it automatically applies these transformations:

Input	Stored as
`<br>`	`<br />`
`<a href="...">`	`<a href="..." rel="ugc noopener noreferrer" target="_blank">`

If you compare the raw value from /overview against the value you originally posted, they will not match exactly. Normalise both sides before comparing.

publishOn must be a Unix timestamp integer — not an ISO 8601 string¶

Passing an ISO 8601 string (e.g. "2026-06-01T21:00:00.000Z") as publishOn in the update body will silently corrupt the value. The server parses only the leading digits as an integer — "2026" becomes 2026 seconds from epoch, producing "1970-01-01T00:33:46.000Z".

Always pass publishOn as a Unix timestamp integer:

publishOn = int(datetime(2026, 6, 1, 21, 0, 0, tzinfo=timezone.utc).timestamp())
# → 1748732400

wizardDraftedToPublishOn behaves oppositely — it requires an ISO 8601 string and returns HTTP 400 if given an integer. See section 1-C for the full example.

Media Upload Flow (Audio / Video)¶

See Recipes — Media upload for the full end-to-end Python example using endpoints 6-A through 6-D.

Supported formats¶

mp3, m4a, aifc, aiff, ogg, wav, flac, mp4, mov

File size limits¶

Type	Limit
Video	4 GB
Audio	500 MB

6-A. Get a Signed Upload URL¶

GET /v3/episodes/{ANCHOR_ID}/upload/signedUrl?filename={FILENAME}&type={MIME_TYPE}&isMumsCompatible=true

Parameter	Example	Description
`filename`	`episode.mp4`	File name to upload
`type`	`video/mp4`	MIME type

Response — two variants depending on storage backend:

Audio files are routed to AWS S3 and return:

{
  "requestUuid": "b8c96d9c-1d37-45a5-7ba3-2dcfb17e1fdc",
  "fileKey": "b8c96d9c-1d37-45a5-7ba3-2dcfb17e1fdc/episode.mp3",
  "signedUrl": "https://audio-uploaded-default-production.s3.us-west-1.amazonaws.com/..."
}

Video files are routed to Google Cloud Storage and return:

{
  "uploadId": "9e1fdfce-4485-6bde-e582-3b7c84978d55",
  "url": "https://storage.googleapis.com/anchor_exclusive_media_upload_production/...",
  "headers": {}
}

Use signedUrl / requestUuid for audio; url / uploadId for video.

6-B. PUT the File to Storage¶

PUT {signedUrl}           # audio (S3)
PUT {url}                 # video (GCS)
Content-Type: {MIME_TYPE}

Save the ETag response header — you need it in the next step.

6-C. Notify Upload Completion¶

POST /v3/upload/{requestUuid}/process_upload?isMumsCompatible=true   # audio
POST /v3/upload/{uploadId}/process_upload?isMumsCompatible=true      # video

Request body (audio):

{
  "userId": YOUR_USER_ID,
  "uploadType": "default",
  "origin": "episode-media:upload",
  "caption": "episode.mp3",
  "isExtractedFromVideo": false,
  "isMultipartUpload": true,
  "parts": [{ "partNumber": 1, "etag": "9e3290d363f9552ada7cb4fa412a9fc7" }],
  "uploadId": "b8c96d9c-1d37-45a5-7ba3-2dcfb17e1fdc",
  "episodeId": 120525547,
  "stationId": YOUR_STATION_ID
}

Request body (video):

{
  "userId": YOUR_USER_ID,
  "uploadType": "video",
  "origin": "episode-media:upload",
  "caption": "episode.mp4",
  "isExtractedFromVideo": true,
  "isMultipartUpload": true,
  "parts": [{ "partNumber": 1, "etag": "4d49fffd0ed1c6a291577ac47d26b997" }],
  "uploadId": "9e1fdfce-4485-6bde-e582-3b7c84978d55",
  "episodeId": 120525547,
  "stationId": YOUR_STATION_ID
}

Parameter	Type	Description
`uploadType`	string	`"default"` for audio; `"video"` for video (not `"audio"`)
`isExtractedFromVideo`	boolean	`true` for mp4/mov; `false` for audio-only
`parts`	array	List of ETags from the storage PUT response
`uploadId`	string	For audio: use `requestUuid` from 6-A. For video: use `uploadId` from 6-A

6-D. Poll for Media Validation¶

GET /v3/upload/media/{requestUuid}?includeMediaValidation=true&isMumsCompatible=true  # audio
GET /v3/upload/media/{uploadId}?includeMediaValidation=true&isMumsCompatible=true     # video

Response field	Description
`status`	`"processing"` / `"completed"` / `"failed"`
`mediaType`	`"video"` / `"audio"`
`durationInMilliseconds`	Duration in ms
`validationStatus`	Validation result

Poll until status is "completed".

Audio polling may return 404 transiently

For audio (S3) uploads, the poll endpoint sometimes returns HTTP 404 for 30–60 seconds after process_upload completes before the record becomes available. Retry with backoff up to ~300 seconds.