Borgcloud
Multilingual speech-to-text API at a flat $0.06/hour with a free tier and $10 signup credit.
Last verified:
What is Borgcloud?
Borgcloud's Speech to Text service transcribes audio files into text through a web interface and a RESTful API. It supports MP3, OPUS, and M4A files with segment-level timestamps. It is aimed at podcasters, students, marketers, and developers who need affordable transcription.
Borgcloud pricing
Pricing model: Freemium
Free tier for experimentation + $0.06/hour pay-as-you-go; $10 free credit for new accounts (about 150 hours)
Borgcloud pros
- Low pay-as-you-go price of $0.06 per hour with no subscriptions or hidden fees
- Timestamp support for every transcription segment
- Multilingual transcription across dozens of languages
- Free tier for experimentation that requires no registration
Borgcloud cons
- Web interface limited to files under 25MB; API accepts files up to 190MB
- Accepts only MP3, OPUS, and M4A formats
- Stated accuracy of over 90% applies to clear English audio, and accents and audio quality can reduce it
Frequently asked questions about Borgcloud
What audio formats are supported?
MP3 (audio/mpeg), OPUS (audio/ogg), and M4A (audio/mp4). Files should be under 25MB for the web interface and up to 190MB via the API.
How is pricing calculated?
$0.06 per hour of audio processed, so a 30-minute file costs $0.03. New users receive $10 in free credit after adding a payment method.
Is there a free tier?
Yes, a free tier for experimentation that requires no registration. It is used by sending Bearer null in API requests. Paid accounts get higher usage limits and priority processing.
How accurate is the transcription?
The service reports over 90% accuracy for clear English audio. Accuracy may vary with accents and audio quality.
What languages are supported?
Dozens of languages, including English, Spanish, French, German, Chinese, Japanese, Arabic, Hindi, and Russian.
Is there an API?
Yes, a RESTful API that accepts files up to 190MB at the same $0.06 per hour rate. Requests go to the audio transcriptions endpoint and can return verbose_json output with segment timestamps.