Create a podcast source-to-broadcast concept pack

Use the generation tools as checkpoints, not as an automatic broadcast pipeline

Start with list_models and get_credits so the selected workflow and available balance are visible before a paid request. Use generate for a text-led cover, thumbnail, or motion concept. For image-to-video, reserve the generation and use upload_input with a rights-cleared still before submission. Track the returned generation identifier with get_generation or wait_generation, retrieve a completed result through download_output, and use cancel_generation only while the documented cancellation path still applies.

Maintain one production row per deliverable. Record the source passage, intended placement, model identifier, ratio, prompt, generation identifier, downloaded file, reviewer, and decision. The package README describes download_output as returning a short-lived signed URL rather than streaming media into the model context, so retrieval is a handoff into the team's controlled workspace, not evidence that the asset has been edited or published. A failed or rejected row should not silently become the source for the remaining pack.

Square cover concept

Text-to-image, one approved visual premiseLeave final title lettering editable in the finishing file and test legibility at small sizes.

Video thumbnail concept

Text-to-image, separate wide compositionDo not crop the square proof and assume the focal subject, safe area, and title space survived.

Short clip concept

Text-to-video or image-to-videoTreat the result as a motion shot for an editor, not as a cut, captioned, mixed, or published episode clip.

Approve a proof budget before expanding the pack

Current first-party catalog charges are model-specific: Recraft V3 is listed at 5 credits for text-to-image brand graphics; Kling 2.6 Pro at 42 credits for text-to-video social video; Kling 2.6 Pro Motion at 42 credits for image-to-video animation; and Veo 3.1 Fast at 96 credits for text-to-video premium campaigns. These are current Studio catalog facts, not a quote for a complete podcast campaign and not evidence of output quality or audience performance.

Use those figures to set a stop rule. Approve one representative still proof first, then one motion proof only if motion is actually required and the source rights are resolved. Review the account's live model list and credit balance before each request rather than multiplying remembered prices into a speculative package total. Do not spend the remaining allowance merely to fill every row: stop when a proof fails source accuracy, usable composition, rights review, or the editor's handoff requirements.

Still proof gate

One model, one ratio, one pass conditionConfirm message accuracy, subject treatment, copy-safe space, and rights status before requesting variants.

Motion proof gate

One shot with an explicit editor handoffApprove the motion purpose and budget separately from the square and wide still concepts.

Match each podcast placement to its real format contract

Square support is concrete but placement-specific. The published MCP generate schema accepts 1:1, 9:16, and 16:9 and defaults to 16:9. Apple Podcasts requires a square Show Cover and currently specifies 3000 by 3000 pixels, while accepting 1400 by 1400 through 3000 by 3000 for RSS submissions; it also requires PNG or JPG with no transparency. YouTube's creator guide also asks for a square podcast thumbnail and defines podcast episodes as videos in the show's playlist.

Video-thumbnail requirements are not interchangeable with podcast-cover requirements. Spotify says a custom video thumbnail is horizontal at 16:9 with a suggested 1920 by 1080 resolution, and distinguishes it from the artwork shown for audio episodes. Generate a dedicated wide composition when that destination matters. The MCP ratio is only the generation frame; the team must still export the correct dimensions, file type, safe area, final text, and platform metadata in downstream tools.

The Studio template catalog demonstrates discrete square brand-poster, text-to-video social-hook, and image-to-video motion recipes. It does not document a single tool that assembles a podcast cover, carousel, thumbnail, and edited clip as one consistent package. Use the same approved subject, palette, lighting direction, and exclusions across rows, but review each result independently instead of assuming separate model calls retain a shared visual memory.

Show or episode cover

Plan a 1:1 compositionFinish exact platform dimensions, solid background requirements, title legibility, and export settings after generation.

Video thumbnail

Plan a 16:9 compositionReserve a focal area and editable title zone; validate the current destination specification before upload.

Carousel

Storyboard outside the generatorChoose approved quotations and slide order first, then create only the visual ingredients each slide needs.

Clip

Generate one bounded motion conceptReturn it to an editor for selection, timing, captions, audio, graphics, and platform-specific export.

Choose the next route from the unresolved production decision

Return to the creator-use-case hub when the source type, production cadence, or required outputs are still undecided. Move to the YouTube creator route when the next task is a thumbnail or short motion concept for a video-first episode. Use the newsletter route when the approved episode source is being translated into a wide editorial hero and social derivative rather than a podcast-platform package. These links separate adjacent production decisions instead of broadening this page into a general content-repurposing guide.

Keep this page within the podcast concept-pack boundary

This page owns the podcast AI asset-pack workflow: an approved episode source, a rights record, distinct square and wide still concepts, a carousel storyboard, one bounded motion concept, a proof budget, and downstream editorial handoff. It does not own a general MCP use-case directory, a YouTube production guide, newsletter repurposing, audio post-production, transcription, social scheduling, or platform publishing. Editors should consolidate it into the parent creator-use-case hub if current evidence can no longer support both square and video concept routes plus source-rights review.

This draft does not use community evidence to support OfflineCreator podcast adoption or outcomes. It makes no claim about creator demand, time saved, consistency, reliability, audience growth, click-through rate, downloads, revenue, or platform acceptance. Product statements are restricted to the current catalog and version-pinned package surface; platform statements are restricted to the current primary guidance retrieved for the named placement.