12 eps × 90 min — what plan survives?
52 Sundays — what does the year cost?
YouTube chapters in one paste
MCP weaverclip_bulk_create_edits — copy into description.
Before you record lesson one
When does your drive fill?
Your 2TB drive, your schedule, your bitrate — one date.
2.70 GB/h · 32.4 GB/week
Paste your Output settings
We read VBitrate and tell you what a night costs.
24–72h — will you lose the best hour?
Podcasts: find the 20 moments worth clipping, not the 2 hours of umms.
Upload the full pod, get a word-timed transcript, rank up to 20 complete thoughts per hour (max 200), render captioned verticals. Your agent can do it overnight and leave a verified review. Free 10 GB/2h holds one 90-min ep, Pro 8h holds a week.
Podcasts are long and repetitive — a 90-minute interview has 12 minutes of shorts hidden in it. Weaver finds the complete thoughts that survive without context: bold takes, stories, lived experience. Not the loudest second, the thought that stands alone. That's why biggest podcasters care: they don't have 90 minutes to scrub, they have 10 seconds to decide which 45s to post. The 20/h ranked inbox is that decision, with timecodes you can feed to weaverclip_bulk_create_edits without downloading video.
Dual 16:9+9:16 sharing one transcript — 90 min is 90 min, no
Full-stream SRT/VTT download to make clips later — upload on
MCP search 'pricing' → 12 hits with startMs/endMs in 300ms,
Bulk 18 edits in one call — 12 hits → 12 queued renders, one
Sharp nuances you didn't know you needed — until you're live
The 90-minute trap — why podcasts feel longer than streams
A 90-minute podcast feels longer than a 2-hour stream because there's no chat to carry dead air. Every minute is you talking, but only 8–12 minutes are distinct, shareable thoughts. Weaver's ranker looks for semantic completeness (a thought with a turn, a payoff), not volume. A 90-minute at 6 Mbps is ~4 GB; 12 eps is ~48 GB and 18h of transcript. Free 2h holds one ep, Creator 4h holds two, Pro 8h holds four, Studio 70h holds the season. The metrics strip shows Storage/Transcript/Renders live so you see '6.2h left' before you pick the next 90-min.
Season planner — the spreadsheet before you record
Podcast Season Planner on /features/transcription: episodes × length × shorts per ep → transcript hrs, renders, GB, 180-day survival. 12 eps × 90 min × 3 shorts = 18h transcript, 36 renders, ~48 GB. Free gives 2h, so you need Pro 8h for month one and Studio 70h for the quarter. The nuance: transcript hours are scarcer than storage, so you don't store 18h and transcribe 18h on day one. Upload 18h, schedule transcribe for the 4 best eps (6h), download full SRT for all 18h (free to stitch), and make clips later. That's the 'plus the clips later' you asked for.
Hook Finder — paste any transcript, see the 10 moments before you sign up
Paste your transcript into Hook Finder on /features/transcription — it mirrors the ranker offline, scores completeness, and returns 10 candidates with in/out timecodes. An agent calls the same logic via transcript text, then weaverclip_bulk_create_edits queues them. For a podcaster, that means you can test the ranker on yesterday's transcript without uploading yesterday's video. If the 10 hits feel like your show, you upload.
Agents overnight — the use case that scales
An agent watches your RSS, ingests the week's raw via /upload-video (browser) or helper import, calls weaverclip_schedule_transcribe for the 90 min, polls manifest until transcript length hits ~80 cues, searches 'sponsor read' to exclude, bulk-creates 9 verticals with Space preset, and leaves a verified review: '12 eps → 9 verticals, 98% verified, host stayed at 14% disk because tmp deleted after verify.' You wake to a gallery link, not a timeline. That's why agents get the same Free 10 GB/2h — an agent that does 3 pods a week never leaves Free, but its reviews build your SEO.
Questions that decide the purchase — answered before you ask
Do I need to transcribe the whole 90 min?
No. Use weaverclip_schedule_transcribe with startMs/endMs. If the intro is 10 min of housekeeping, skip it. Transcribe 00:10:00 to 01:20:00 — 70 min, not 90, saving 20 min of transcript hours every ep. Over 12 eps that's 4h saved — a full Creator month.
Can I get the full transcript as SRT to archive?
Yes. Every session with transcript shows FullStreamCaptions on /sessions/[id] — SRT and VTT stitched from transcript_chunks, same chunks that power 20/h. Download tonight, import to Descript, make clips next week without re-transcribing. For a podcaster, that's the LMS archive: 90 min SRT is ~120 cues, ~15 KB.
What about dual video — my face + screen share?
Record both canvases via OBS + Aitum vertical, same filename stem, helper pairs on exact stem, both masters count as storage (2× bytes) but share one transcript (counts once). Each finished export counts as one render — vertical Short + square feed = 2 renders from same 90 min.
How do I not lose the best moment in a 90 min?
You don't watch 90 min. You search. weaverclip_search_transcript query 'burnout' across 90 min returns 4 hits with startMs/endMs in 300ms, no video download. Or open the Safety Rail — 90 min is 88 ticks, gaps are red, verified is green — and jump to the densest ticks where you talked fastest. The ranked inbox is already sorted by score, not chronology.
Try it with the video you already have.
Free holds a full 2–3h webinar at 6 Mbps for 30 days. Upload the whole thing, download the SRT, make clips later — no re-transcribe. Every plan shows the numbers before you pay.