Add-on: Transcription Service
$79/year — Bring your own API key
Transcripts at near-wholesale API cost, orchestrated by Benecaster.
Get Transcription Service — $79/year
AI transcription without a middleman markup.
A transcript is a different kind of audience. Some listeners read instead of listen — deaf and hard-of-hearing audiences, non-native speakers, people who are skimming in a situation where audio doesn’t work. Transcripts open your episodes to all of them.
They also do something audio can’t: get indexed by search engines. The guest’s name, the niche topic, the specific tool mentioned at 34 minutes — none of that is visible to Google without a transcript. With one, the actual conversation becomes part of your SEO footprint.
And since Benecaster controls transcript visibility per tier, a transcript can be a concrete subscriber incentive. Offer transcript access as a premium-tier benefit. If you choose, show new visitors that transcripts exist before they sign up, with a notice or the first few lines. Existing subscribers get a reason to stay on the paid tier; prospective ones get a preview of what they’re missing.
Most podcast operators haven’t bothered — because the workflow is genuinely annoying. Either you’re paying a transcription service that marks up the underlying API cost, or you’re managing API calls and file storage yourself.
Transcription Service removes that friction. It connects Benecaster to your own transcription API account — OpenAI Whisper, AssemblyAI, Deepgram, or Rev.ai. You bring the API key; transcription costs go directly to your provider at their published rates. Benecaster doesn’t mark them up or sit between you and the billing. Request a transcription from the episode editor or the episode list. Benecaster confirms before anything is charged, the job runs in the background, and the transcript appears when it’s done.
The transcript that comes back isn’t a text file in a folder. It’s timestamped, editable, and displayed as an interactive reader — listeners click any timestamp to jump the player to that moment. The full text is indexed for search, so subscribers can find the episode where something was discussed, with the words around it highlighted. You choose which episodes get transcribed and which don’t.
Four providers. You choose.
| Provider | Labels who’s speaking (diarization) | Notes |
|---|---|---|
| OpenAI Whisper | Only with OpenAI’s gpt-4o-transcribe-diarize model, not the original Whisper model | Popular, well-documented, competitive per-minute pricing. Takes audio files up to 25 MB, about 25 minutes at a typical podcast bitrate, so it suits shorter episodes |
| AssemblyAI | Yes | Strong accuracy |
| Deepgram | Yes | Fast, cost-effective at scale |
| Rev.ai | Yes | Human review options available via their platform for high-accuracy needs |
OpenAI receives the audio file itself and refuses anything over 25 MB. The other three fetch your episode from its URL, with no size limit. For episodes longer than about 25 minutes, choose AssemblyAI, Deepgram or Rev.ai. If an episode is too large for OpenAI, the job stops before anything is sent or charged.
Per-show and per-episode transcription is opt-in independently — enable transcription for the shows and episodes where it matters.
Transcripts that do more than sit on a page.
- Timestamped transcript storage with the full provider response
- Interactive transcript on episode pages — click any timestamp to jump the audio player
- Plain text extraction for search indexing (works with built-in Standard Search and AI Search add-on)
- Choose who reads each show’s transcripts: all visitors, all subscribers, selected tiers, buy-up holders, or nobody while you review — and override it for any episode
- Choose what everyone else sees, per show: nothing, a notice with a sign-up link, or a notice and the first few lines
- The transcript visibility setting applies to the on-page transcript, site search and the
<podcast:transcript>tag in each listener’s feed — listeners it excludes don’t see the transcript in their podcast app either - SEO-optimized transcript rendering
- Transcript edit and correction interface — fix mishearings before publishing
- Export transcript as SRT (video captions), VTT (web players), plain text, or PDF (Latin, Greek and Cyrillic letters only; other writing systems export as SRT, VTT and plain text)
- Let listeners who may read a transcript download it too, under the on-page transcript
- Import transcripts you already have: SRT, WebVTT, plain text or OpenAI Whisper JSON, by upload, paste or from your old host’s RSS feed. Free, with no provider call
- Format transcript: one click has OpenAI lay it out in paragraphs with speaker names in bold, for the page, the PDF and podcast apps that read HTML transcripts. Timestamps stay clickable. Charged to your own OpenAI account, typically under $0.01 an episode; with AssemblyAI, Deepgram or Rev.ai it needs a separate OpenAI key
- Episode list status indicator: each episode shows its transcription state — not configured, ready to transcribe, queued, processing, complete, or failed
- Request a transcription from the episode list; Benecaster confirms before anything is charged to your account, and the job runs in the background
- Transcribe a whole back catalogue in one go, with the estimated cost shown before anything is queued, and pause the queue whenever you like
- See an estimate of what you’ve transcribed this month, right on the settings page
- Label who’s speaking, per show, in the transcript and in every export
- A transcript for each tier’s own audio: each subscriber’s feed carries the transcript of what they hear
What does transcription actually cost?
Transcription costs vary by provider and audio length. As a rough guide, from each provider’s list price as of November 2026:
- OpenAI Whisper: approximately $0.006/minute — a 20-minute episode costs ~$0.12 (OpenAI takes files up to 25 MB, so it suits shorter episodes)
- AssemblyAI: up to about $0.004/minute, depending on the model and speaker labels
- Deepgram: about $0.0043/minute
- Rev.ai: up to about $0.005/minute, depending on the language
A podcaster releasing two 60-minute episodes per month spends under $1/month in API costs at these rates. Provider pricing changes — check each provider’s current rates before selecting.
Benecaster adds no markup. The $79/year is for the add-on software; API usage is billed directly by your provider to your account.
Even More Powerful With
- AI Search — Transcript content is indexed alongside episode titles and show notes, so subscribers can search the spoken content of your episodes in plain English. For interview shows, this is the difference between AI Search finding the right answer and missing it entirely — most of what was actually said lives in the audio, not the description.
- Guest Manager — When diarization is enabled, transcript segments are attributed to speakers. Map those speaker labels to Guest Manager profiles and the connection runs both ways: guests accumulate attributed transcripts on their archive pages; you can search your back catalog by what a specific guest said.
- Standard Search (built-in) — Transcript text is automatically included in keyword search results with no additional configuration.
- Blocks & Widgets Library — Place a transcript on any page as a block or an Elementor widget, with no shortcode to type: in full, as a short preview, or in a box that scrolls and follows the player.
When all three are active together — Transcription, Guest Manager, and AI Search — your guest content becomes fully attributed and semantically searchable. Subscribers can find what a specific guest said about a specific topic, across your entire archive.
Requires / Works with
- Requires: Benecaster, on any plan from Launch up
- Requires: An API key from a supported transcription provider (OpenAI, AssemblyAI, Deepgram, or Rev.ai)
- Works with: Built-in Standard Search (included free) and AI Search add-on for transcript-inclusive episode search