Groq's Speech to Text API runs OpenAI Whisper models on Groq LPUs through OpenAI-compatible file endpoints. Public list prices on the Console docs (checked October 25, 2026) are $0.111 per hour for whisper-large-v3 and $0.04 per hour for whisper-large-v3-turbo.
The LecSync Realtime Speech API is a managed WebSocket for meeting and lecture captions: preview and final text, translation between any two of 60 languages, a speaker and audio-aligned timestamps on every final segment, and running minutes. It costs $0.006 per audio minute ($0.36 per hour) with those features included.
Figures below were checked on October 25, 2026 against Groq's Speech to Text docs (also served as markdown), the whisper-large-v3 and whisper-large-v3-turbo model pages, and rate limits. The marketing URL groq.com/pricing no longer lists STT rates; dollar figures come from Console documentation only.
What is Groq Whisper vs a realtime STT caption API?
Groq Whisper is a high-throughput batch-style speech-to-text service: you upload an audio file or pass a URL to /openai/v1/audio/transcriptions (or /translations for English-only output). A realtime STT caption API such as LecSync keeps a WebSocket open, streams PCM as people talk, and returns paired transcripts, translations, speakers and minutes without waiting for a finished file.
The short answer
If you need cheap, fast file transcription (or English-only Whisper translation) and can assemble your own live pipeline, Groq Whisper undercuts almost every managed caption bill. Turbo lists at $0.04 per hour of audio; large-v3 at $0.111 per hour — far below LecSync's $0.36 per hour all-in.
If you need live translated captions, speakers and minutes on one flat minute bill while the session is still running, LecSync is the closer product. Groq's public STT path is file/URL based; it does not document a managed WebSocket that emits preview→final captions with live text translation across arbitrary language pairs or live speaker labels.
How each API works
LecSync. Your server calls POST https://api.lecsync.com/v1/realtime/connect with the audio format, a language pair and the features you want, then streams 16 kHz mono PCM. Preview text arrives while someone is still talking; a final segment settles once the sentence confirms. Translations share the segment ID. With speakers: true, each final carries a speaker and startMs/endMs on the recording clock. With digest: true, minutes arrive section by section. The first word shows up about 0.28 seconds after speech starts; translation adds roughly 88 ms. Seven gateway regions (dal, jp3, kc, nj, ru2, ru3, uk) keep audio near the capture point; backlog is capped at 2 seconds.
Groq Whisper. You send file or url to the transcriptions endpoint with model set to whisper-large-v3 or whisper-large-v3-turbo. Optional language, prompt, temperature, and timestamp_granularities (word / segment with verbose_json) shape the result. The translations endpoint turns speech into English text only and is supported on large-v3, not turbo. Files under 10 seconds are still billed for 10 seconds. Free tier max upload is 25 MB (100 MB on the dev tier); larger audio needs chunking (Groq Cookbook) or a URL. Docs quote real-time speed factors of 189× (v3) and 216× (turbo) — throughput relative to audio duration, not first-word caption latency on a live socket.
Pricing converted to dollars per hour (Estimated)
| Path | Published rate | Implied $/hour of audio | Notes |
|---|---|---|---|
| LecSync Realtime Speech API | $0.006 / min all-in | $0.36 | STT + translation + speakers + digest included |
Groq whisper-large-v3-turbo | $0.04 / hour | $0.04 | Transcription only; no translation per docs table |
Groq whisper-large-v3 | $0.111 / hour | $0.111 | Transcription + English translation endpoint |
Minute equivalents (Estimated): turbo ≈ $0.00067/min; v3 ≈ $0.00185/min. LecSync stays $0.006/min whether translation and speakers are on or off.
Side-by-side facts (checked October 25, 2026)
| LecSync Realtime Speech API | Groq Whisper STT | |
|---|---|---|
| Primary job | Live meeting/lecture captions + MT | Fast file transcription (+ EN translation on v3) |
| Price model | $0.006/min flat, all features | $0.04 or $0.111 per audio hour |
| Delivery | Managed WebSocket, preview→final | HTTP file/URL upload |
| Live text translation | Any two of 60, same socket | Translations endpoint → English only (v3); turbo: no |
| Speakers on live stream | Every final segment | Not a documented STT feature |
| Live minutes / digest | digest: true | Not available |
| Timestamps | Recording-aligned startMs/endMs | Word/segment via verbose_json |
| Languages | 60, auto-detected | 99+ (model cards) |
| Latency story | ~0.28 s first word; +~88 ms MT | Speed factor 189–216× (batch throughput) |
| Min billing | Fractional seconds; empty stream not billed | 10-second minimum per request |
| Hosting | Managed SaaS, 7 gateway regions | Groq cloud inference |
| Rate limits (docs) | — | 20 RPM; ASH 7.2K; ASD 28.8K per Whisper model |
What 1,000 hours cost (Estimated)
| Workload, 1,000 hours a month | LecSync | Groq (from public $/hr) |
|---|---|---|
| Same-language captions / transcripts | $360 | Turbo $40; v3 $111 |
| Translated live captions (arbitrary pair) | $360 | Not on one STT socket; DIY STT + MT |
| Speakers with timestamps during the session | Included | Not documented on Whisper STT |
| Live minutes | Included | Not available |
| Post-call English translation of a file | Possible via your own stack | v3 translations endpoint (EN only) |
Groq wins hard on raw STT dollars for finished audio. LecSync wins when the product is a live bilingual caption feed rather than a file job. For a deeper flat-vs-add-on walkthrough, see what 1,000 hours cost once translation, speakers and summaries are added. For stitching your own STT + MT, see built-in translation vs a DIY pipeline.
When Groq Whisper is the better pick
- Batch or post-call transcription of files/URLs at $0.04–$0.111 per hour.
- Maximum language coverage on STT alone (99+ on the model cards vs LecSync's 60 with built-in MT).
- English-only speech translation via the dedicated translations endpoint on large-v3.
- OpenAI-compatible SDK drop-in if you already call Whisper-style HTTP APIs.
- High throughput jobs where 189–216× real-time speed factors matter more than first-word socket latency.
When LecSync fits better
- Translated live captions on one socket — any pair of 60 languages, each translation tied to its segment.
- Speakers and recording-aligned timestamps during the live stream — meeting client with speakers and timestamps.
- Live minutes while the session runs (
digest: true) — live digest vs post-call summarization. - Predictable minute billing at $0.36 an hour with no separate translation SKU and no 10-second floor per request.
- Classroom / study-abroad style feeds where captions must keep up as the lecturer switches languages, not after the recording uploads.
FAQ
Is Groq Whisper a realtime streaming caption API?
Not in the sense LecSync is. Public docs expose file/URL transcription and translation HTTP endpoints. Model pages describe high real-time speed factors for how fast a file finishes relative to its duration. They do not document a managed WebSocket that streams preview and final captions with live multilingual text translation.
Why is LecSync more expensive per hour of audio?
LecSync's $0.36/hour includes live translation, speakers, glossary/polish and digest on the same connection. Groq's $0.04–$0.111/hour buys Whisper transcription (and English translation on v3). Building live bilingual captions on Groq means adding your own streaming front end, MT, diarization and minutes — and owning that glue.
Does Groq return speaker labels?
The Speech to Text docs do not offer a speaker-diarization parameter. They mention overlapping speakers as a condition that can hurt confidence scores. For live speaker-tagged segments, LecSync exposes speakers: true on the realtime socket.
How do the 10-second minimum and file size limits affect cost?
Every Groq STT request bills at least 10 seconds, even for shorter clips. Free tier uploads stop at 25 MB (100 MB on the dev tier). Short interactive turns or very long recordings need different packaging than LecSync's continuous stream with fractional-second billing.
Can I mix both?
Yes. Many teams run LecSync for the live classroom or meeting feed, then push archived audio to Groq Whisper for cheap re-transcription or English-only offline translation. The APIs solve different layers of the same product.
Next step
Open the LecSync Realtime Speech API docs, create a connect session, and stream three seconds of silence to see preview and final segments on one socket — translation, speakers and digest included at $0.006 per audio minute.