# Groq speech to text

Hosted Whisper speech-to-text API that transcribes audio files, or translates them to English, with OpenAI-compatible endpoints.

- Provider: Groq (https://groq.com)
- Category: Transcribe audio
- Type: Hosted API
- Pricing model: Pay as you go on the Developer tier, billed per hour of audio [3]
- Free tier: Free tier with lower limits (25 MB max file size, 20 requests per minute) [1]
- Authentication: API key sent as a bearer token [6]
- Connect via: REST API (OpenAI-compatible), Python and JavaScript SDKs [1]
- Read from the provider's pages on 2026-11-04
- Page: https://supermarket.dev/tools/transcribe-audio/groq-speech-to-text

## Pricing

### Billing
- Pricing model: Pay as you go, invoiced monthly in arrears [3]
- Free tier: Free tier exists with lower limits (25 MB max file size) [1]
- Card required for free tier: Not published
- Minimum spend: Not published
- Billing period: Monthly; early invoices when cumulative usage reaches $1, $10, $100, $500 and $1,000 [3]

### Prices
Prices are per hour of audio as published, no conversion.
| Model | $ per hour of audio | Notes |
|---|---|---|
| `whisper-large-v3` | $0.111 | Transcription and translation |
| `whisper-large-v3-turbo` | $0.04 | Transcription only |
Source [1]

### Other charges
- Minimum billed length per request: 10 seconds [1]
- Card to upgrade to Developer tier: Valid payment method needed (credit card, US bank account or SEPA debit) [3]

## Tools

### MCP server
- Official MCP server: No [13]
- Address or package: Not published

### API
| Endpoint | What it does |
|---|---|
| `POST https://api.groq.com/openai/v1/audio/transcriptions` | Convert audio to text |
| `POST https://api.groq.com/openai/v1/audio/translations` | Translate audio to English text |
Source [1]

### Features
- Default model: No default; `model` is required (`whisper-large-v3-turbo` in the docs example) [1]
- Languages: Multilingual, 99+ languages; translation output is English only [2]
- Speaker labels: Not published
- Word timestamps: Yes, with `response_format` `verbose_json` and `timestamp_granularities` `word` or `segment` [1]
- Live streaming: Not a streaming API; files or URLs are submitted per request [1]
- Max file size: 25 MB free tier, 100 MB dev tier; direct upload 25 MB, larger via `url` parameter [1]
- Audio input: `flac`, `mp3`, `mp4`, `mpeg`, `mpga`, `m4a`, `ogg`, `wav`, `webm`; file upload or URL; first track only [1]

## Authentication

### Methods
| Method | Used for | How |
|---|---|---|
| API key | All API calls | `Authorization: Bearer $GROQ_API_KEY` header |
Source [6]

### Getting access
- Who creates the account: Not published
- Where to get a key: Set the key as `GROQ_API_KEY` [7]
- Other requirements: Developer tier needs a payment method [3]

## Setup

### Addresses
- REST API: `https://api.groq.com/openai/v1/audio/transcriptions` [1]
- MCP server: Not published
- OpenAPI spec: Not published
- Docs for agents: `https://console.groq.com/llms.txt` [9]
- Status page: `https://status.groq.com` [13]

### Install
| What | Command |
|---|---|
| Python SDK | `pip install groq` |
| JavaScript SDK | `npm install --save groq-sdk` |
Source [8]

## Limits

### Rate limits
Base Developer plan limits for both Whisper models. ASH is audio seconds per hour, ASD per day.
| Model | RPM | RPD | ASH | ASD |
|---|---|---|---|---|
| `whisper-large-v3` | 20 | 2K | 7.2K | 28.8K |
| `whisper-large-v3-turbo` | 20 | 2K | 7.2K | 28.8K |
Source [4]

### Other limits
- Over the limit: HTTP 429 with a `retry-after` header [4]
- Prompt length: 224 tokens [1]
- Minimum file length: 0.01 seconds [1]
- Scope of limits: Per organization [4]

## Privacy

### Your data
- Data location: United States (Google Cloud Platform buckets) [5]
- Retention: Not retained by default for inference; logs kept up to 30 days only for troubleshooting or abuse investigation [5]
- Used for training: Not published
- Zero data retention: Available to all customers in Data Controls settings; covers transcription endpoints [5]

### Compliance
- Certifications: Not published
- HIPAA: Not published

## Sources

1. Speech to text: https://console.groq.com/docs/speech-to-text.md
2. Whisper Large V3 Turbo model page: https://console.groq.com/docs/model/whisper-large-v3-turbo.md
3. Billing FAQs: https://console.groq.com/docs/billing-faqs.md
4. Rate limits: https://console.groq.com/docs/rate-limits.md
5. Your data in GroqCloud: https://console.groq.com/docs/your-data.md
6. API reference: https://console.groq.com/docs/api-reference.md
7. Quickstart: https://console.groq.com/docs/quickstart.md
8. Libraries: https://console.groq.com/docs/libraries.md
9. llms.txt: https://console.groq.com/llms.txt
10. MCP docs: https://console.groq.com/docs/mcp
11. Pricing: https://groq.com/pricing
12. Trust center: https://trust.groq.com/
13. Checked by supermarket.dev