accept_clips
ChatGPTRender the chosen highlights as clips. Spends credits; credits has to equal the current price or the call is refused (QUOTE_CHANGED). The same requestId and expectedRevision on a retry do not charge twice.
add_topic
ChatGPTQueue one typed topic (with optional notes and link) for the series' next videos.
approve_drama_cast
ChatGPTApprove one cast member (person + castId) or the setting. Spends nothing.
clip_highlights
ChatGPTThe highlights found in an uploaded video and their price. Read only.
clip_upload_url
ChatGPTOpen a resumable upload for a long video the user owns. Spends nothing.
complete_clip_upload
ChatGPTTell SeedViral the upload finished so highlights can be found (free). Spends nothing.
configure_music_video
ChatGPTReplace the saved music video settings at expectedRevision. Overwrites the previous settings; spends no credits and does not render a video. Spends nothing.
create_drama
ChatGPTWrite the plan, cast and setting for a short drama (free until make_drama). Spends nothing.
create_series_draft
ChatGPTSave a new series draft. Only settings.topic is needed (for example {"topic": "history"}); voice, art, captions, length and schedule have defaults that can be changed. Free: it does not start a plan, make a video or post anything.
create_song
ChatGPTStart a song from a brief or lyrics; write: true drafts lyrics for free. Spends nothing.
create_video
ChatGPTSpend the credits of a quote from quote_video and queue the video. One quote makes at most one video; a retry returns the same video.
download_video
ChatGPTGet a fresh five-minute download link for a finished video. Does not write a file or change the video.
final_drama_episode
ChatGPTStart a final episode that ends the story. Spends nothing.
get_balance
ChatGPTSpendable credits and this month's plan credits.
get_catalog
ChatGPTReleased video formats, niches, languages, voices, art and caption styles, durations and credit costs. counts holds the current number of each.
get_download_link
ChatGPTA short-lived link to a finished video's MP4.
get_drama
ChatGPTA drama's state, cast, setting and, once ready, a five-minute video link. Read only.
get_music
ChatGPTA song or music video project, its revision and price. Read only.
get_series
ChatGPTOne series: draft settings, state, schedule and plan status.
get_topics
ChatGPTThe series' topic source (RSS feed, YouTube channel, CSV list or typed topics), upcoming topics and history.
get_video
ChatGPTOne video's stage (scheduled, generating, ready, failed), title, topic and timing.
get_voice_change
ChatGPTA voice change's state, price and, once ready, a five-minute audio link. Read only.
get_workspace
ChatGPTWhich SeedViral workspace this key belongs to and its permissions (read, create).
list_exports
ChatGPTRequested shapes of a video and download links for ready ones.
list_presenters
ChatGPTSaved presenters (ids and names). Read only.
list_series
ChatGPTSeries in this workspace, newest activity first.
list_video_versions
ChatGPTEdited, dubbed or rebuilt versions kept beside a video.
list_videos
ChatGPTVideos, newest first; optionally for one series.
make_drama
ChatGPTRender an approved drama for the credits get_drama shows. Spends credits; credits has to equal the current price or the call is refused (QUOTE_CHANGED). The same requestId and expectedRevision on a retry do not charge twice.
make_music_video
ChatGPTRender the music video at the quoted credits. Spends credits; credits has to equal the current price or the call is refused (QUOTE_CHANGED). The same requestId and expectedRevision on a retry do not charge twice.
make_song
ChatGPTSing the song from the saved draft. Spends credits; credits has to equal the current price or the call is refused (QUOTE_CHANGED). The same requestId and expectedRevision on a retry do not charge twice.
next_drama_episode
ChatGPTStart the next episode with the same cast. Spends nothing.
pause_topic_source
ChatGPTPause automatic topic collection from the current source. Existing queued topics and videos are kept.
preview_topic_source
ChatGPTShow which topics a source would queue. Saves nothing. rss/youtube: input is the link (titles and descriptions only are read). csv: input is {fileName, text, mapping?}; the result lists the columns used (mapping), the header row, row errors and a previewHash that use_topic_source needs.
quote_dub
ChatGPTPrice a dub of a finished video into another language. Spends nothing.
quote_edit
ChatGPTPrice an edit of a finished video. Spends nothing.
quote_video
ChatGPTPrice one video. Returns the credits and the frozen input, valid for 15 minutes; nothing is charged. Modes: prompt (an idea, 8+ chars), script (40+ chars, the user's words), link (reads the public article, Reddit or YouTube page), remix (like/part2).
remove_topic_source
ChatGPTRemove the current topic source and its queued topics. Already made videos are kept.
request_export
ChatGPTAsk for a square (1:1) or landscape (16:9) version of a finished video. Free.
requeue_topic
ChatGPTReturn one skipped or failed topic to the series queue.
research_formats
ChatGPTReleased video formats with examples; count is the current number of formats. Read only.
resume_topic_source
ChatGPTResume automatic topic collection from a paused source.
skip_topic
ChatGPTSkip one queued topic so it will not be used for a video.
update_series_draft
ChatGPTChange some settings of a series. Send only the settings that change, with expectedRevision from get_series.
use_topic_source
ChatGPTUse a feed, channel or CSV list for the series. expectedSourceId is the current source id (null if none); a different id is refused. Replacing a source drops its queued topics; videos already made keep theirs. For csv, the input and previewHash come from preview_topic_source.
voice_changer
ChatGPTConvert the user's uploaded speech to the chosen catalog voice, for the credits get_voice_change shows. Spends credits; credits has to equal the current price or the call is refused (QUOTE_CHANGED). The same requestId and expectedRevision on a retry do not charge twice.
voice_upload_url
ChatGPTA five-minute link to upload (PUT) a recording of the user's own English speech, which is converted to the chosen catalog voice. The recording is not used to identify anyone. Spends nothing.