Google Cloud Speech-to-Text
GCP Speech v2 with Chirp 2 foundation model, batch + streaming, 125+ language variants.
GCP Speech v2 with Chirp 2 foundation model, batch + streaming, 125+ language variants.
Best for gCP-native teams, multilingual coverage, anyone wanting access to Google's Chirp foundation models. Pricing: from $0.016/min (v2 standard) / Chirp tiered.
What it is
Google Cloud Speech-to-Text v2 is the current generation of GCP's managed ASR. It includes the legacy phone-call and video models plus the Chirp and Chirp 2 universal speech models, a 2B-parameter family trained on multilingual data. Features include automatic punctuation, word-level timestamps, speaker diarization, profanity filtering, model adaptation (custom phrase hints and classes), and dynamic batching. V2 introduces recognizer resources that pin model + language. Data-logging opt-in gives a price discount. HIPAA-eligible under a signed GCP BAA. Available across most GCP regions; Chirp/Chirp 2 are limited to a subset.
Watch out for: Chirp 2 region availability is narrower than v1; pricing varies by feature flag (data logging discount, model selection).
Install / use
gcloud ml speech recognize ... # or REST v2Features
| Speaker diarization | Yes |
| Word-level timestamps | Yes |
| Streaming / real-time | Yes |
| Languages supported | 125 |
| HIPAA eligible | Yes |
Google Cloud Speech-to-Text vs Whipscribe
| Feature | Google Cloud Speech-to-Text | Whipscribe |
|---|---|---|
| Category | Transcription APIs | Transcription APIs |
| Pricing | from $0.016/min (v2 standard) / Chirp tiered | free beta |
| Speaker diarization | Yes | Yes |
| Word timestamps | Yes | Yes |
| Streaming | Yes | No |
| Languages | 125 | 99 |
| Platforms | API | Web, API, MCP |
Alternatives to Google Cloud Speech-to-Text
Whipscribe is a managed faster-whisper + whisperX service. If you want transcripts without running infrastructure, paste a URL or drop a file in the form below — you'll have a transcript in seconds.