Vendor page
20x faster Whisper than OpenAI - Fireworks audio transcribes 1 hour in 4 seconds
Fireworks AI, Fireworks AI, 2024
Read the originalRead on October 19, 2026
- What it is
- Fireworks AI's launch post for its audio transcription API, published December 9, 2024. It describes serverless audio APIs billed per audio minute and gives a table of async transcription prices per audio minute for Whisper v3 large and Whisper v3 large turbo, next to latency and throughput.
- How much weight it carries
- Written by the vendor, so it states the company's own prices, but it is a 2024 beta-era post. The page's date metadata was updated in 2026, yet the table is not labeled with a date, and the current Fireworks pricing page lists GPU prices and no audio or Whisper rows. Treat the per-minute figures as the prices announced in 2024, not confirmed current list prices.
Facts taken from this source
- Fireworks AI's audio transcription launch post lists serverless Whisper-v3-large at $0.0015 per audio minute in its async transcription table.
“[Serverless] Whisper-v3-large (beam size 5) $0.0015 11 sec Variable” (Async Transcription Measurements table)
- Fireworks AI's audio transcription launch post lists serverless Whisper-v3-large-turbo at $0.0009 per audio minute in its async transcription table.
“[Serverless] Whisper-v3-large-turbo (beam size 5) $0.0009 7 sec Variable” (Async Transcription Measurements table)