Podcast to multilingual subtitles in one step

Description and use

Apify’s audio‑transcriber Actor turns a podcast into multilingual subtitles in one run, preserving exact timestamps.

Problem

Transcribing a podcast and generating accurate, timestamp‑aligned subtitles in multiple languages typically requires three separate steps: transcription, cue segmentation, and translation, which can break timings and merge cues.

Outcome

For a 4 min 57 s NASA podcast, the Actor produced 87 cues in English, Spanish, and Traditional Chinese with identical timestamps, in 75 s of processing time and $0.05 cost.

Workflow

• Provide podcast RSS feed, episode filter, and list of target languages via JSON input
• Actor fetches episode audio, transcribes with Whisper large‑v3‑turbo
• Actor segments transcript into subtitle‑sized cues (≤7 s, ≤2 lines) using word timings
• Actor translates each cue into requested languages, preserving timestamps
• Actor outputs dataset row with transcript, segments, word timestamps, and SRT/VTT URLs for each language

Evidence

Every cue in the translated files has exactly the same timestamps as the English one.

Source

The run took 75 seconds for transcription and both translations (81 s for the whole run).

Source

Price: 5 audio minutes × $0.006 + 5 minutes × 2 languages × $0.002 = $0.05

Source