Translate Portuguese conversations

[kalany]

What is the best tool (better if is self-hosted or open source) to transcript Portuguese podcasts in .wav files?

Thanks a lot for your help!

[Orionpax]

@kalany For transcribing Portuguese podcasts in .wav files, here is the best self-hosted/open-source options you can integrate with n8n:

Whisper (OpenAI) - Recommended

• Excellent Portuguese support (Brazilian & European)

• Open-source (can self-host)

• High accuracy for podcasts

• Handles .wav files natively

• Free to use locally

Self-hosted setup:

• Run Whisper locally via Docker or Python

• Use n8n’s Execute Command or HTTP Request node to call it

• Models: small, medium, or large (larger = more accurate but slower)

n8n integration:

Read Binary File (.wav)
    ↓
HTTP Request to local Whisper API
    ↓
Get transcription text

Hope this helps!

[Anshul_Namdev]

Hi @kalany What i recommend is try using STT API directly, i mean plug your .wav files into the AWS transcribe node or call a service like wit.ai via the HTTP request node, that would be a better take and a seamless approach:

AWS Transcribe node documentation | n8n Docs

[achamm]

Whisper is what you want here, it handles Portuguese really well and you can self-host it. There’s actually a community node called n8n-nodes-transcribe-audio that runs Whisper locally right inside n8n so you don’t need any external API or server, just install it from the community nodes menu and point it at your .wav files.