Back to Blog

LecSync vs Deepgram Nova-3: Streaming Rates After the Promo

LecSync Team

On Deepgram’s pricing page (checked November 3, 2026), streaming Pay As You Go rates for Flux STT English and Nova-3 Monolingual both list at $0.0077 per audio minute, Flux Multilingual at $0.0078, and Nova-3 Multilingual at $0.0092. Growth prepaid rates are $0.0065 / $0.0068 / $0.0065 / $0.0078 for those four rows. The limited-time promotional figures that appeared on the same page on October 28, 2026 (Flux English $0.0065, Nova-3 Monolingual streaming $0.0048) are not shown in today’s Streaming table.

The LecSync Realtime Speech API is a managed WebSocket for live 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 November 3, 2026 against Deepgram pricing, the models & languages overview, and LecSync developers. This refresh is separate from the earlier LecSync vs Deepgram Nova and LecSync vs Deepgram Flux posts: the public rate card changed, and the product split between Flux (agents) and Nova-3 (meetings/captions) is unchanged.

What is Deepgram Nova-3 / Flux vs a realtime caption API?

Deepgram Nova-3 is a general-purpose ASR SKU for meetings, event captioning, and multi-speaker audio in streaming or pre-recorded mode, with monolingual and multilingual price rows and 60+ languages on the docs overview. Deepgram Flux is a conversational streaming STT SKU with model-native turn detection for voice agents (flux-general-en / flux-general-multi, 10 languages on Multi). A realtime caption API such as LecSync keeps one session open for live bilingual captions, speakers and minutes while people are still talking — including lecture and meeting flows where turn-taking for an LLM is not the product.

The short answer

If you want Deepgram STT alone — Nova-3 for meetings/captions or Flux for agent turn-taking, with optional paid add-ons and separate Audio Intelligence tokens — Deepgram’s public card is clear and bills by the second. Pre-recorded Nova-3 Monolingual at $0.0043/min still undercuts LecSync for file transcripts.

If you need live bilingual captions + speakers + digest on one flat minute bill, LecSync is the closer product. Deepgram streaming STT does not package any-to-any live text translation and live minutes into the same $0.0077–$0.0092 streaming minute; speaker diarization on streaming is a $0.0020/min add-on on Pay As You Go.

How each path 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.

Deepgram Nova-3 / Flux. Connect to Deepgram’s listen APIs (Flux uses the Flux listen path documented for /v2/listen; Nova streams on the Nova listen path). Choose nova-3 for general captioning/meetings or a Flux model for agent turn detection. Enable streaming features as query/params; diarization, keyterm prompting, redaction and entity detection bill as add-ons on top of the base STT minute when you turn them on. Summarization and other Audio Intelligence features bill per 1,000 input/output tokens in addition to transcription on the pricing page. There is no LecSync-style any-to-any live text MT + live digest bundled into the STT streaming rate.

Pricing: Nov 3 card vs Oct 28 promo snapshot

Streaming SKU (Pay As You Go)Oct 28, 2026 (prior post)Nov 3, 2026 (owned page today)
Flux English$0.0065/min promo (list $0.0077)$0.0077 / min
Flux Multilingual$0.0078/min$0.0078 / min (unchanged)
Nova-3 Monolingual$0.0048/min promo (list $0.0077)$0.0077 / min
Nova-3 Multilingual$0.0058/min promo (list $0.0092)$0.0092 / min
LecSync Realtime Speech API$0.006/min all-in$0.006 / min all-in

Growth plan streaming (Nov 3): Flux EN / Nova-3 Mono $0.0065/min; Flux Multi $0.0068; Nova-3 Multi $0.0078. Pre-recorded Nova-3 Mono $0.0043/min, Multi $0.0052/min, Whisper Large $0.0048/min. Streaming Speaker Diarization add-on: $0.0020/min PAYG.

Calculated from Measured $/min × 60 / × 60,000:

Path$/min$/hr1,000 h
LecSync all-in0.0060.36$360
Flux EN or Nova-3 Mono streaming PAYG0.00770.462$462
Flux Multi PAYG0.00780.468$468
Nova-3 Multi streaming PAYG0.00920.552$552
Nova-3 Mono streaming + diarization PAYG0.00970.582$582
Nova-3 Mono pre-recorded PAYG0.00430.258$258

LecSync stays $0.006/min whether translation, speakers and digest are on or off → 1,000 audio hours = $360.

Side-by-side facts (checked November 3, 2026)

LecSync Realtime Speech APIDeepgram Nova-3 / Flux
Primary jobLive meeting/lecture captions + MTSTT (Nova-3 meetings/captions; Flux agents)
Streaming PAYG base$0.006/audio-min flat, all features$0.0077–$0.0092/min by SKU (no promo on page today)
DeliveryManaged WebSocket, preview→finalStreaming listen APIs; Flux turn events
Live any-to-any text MT60 languages, same socketNot in STT streaming SKU; build separately
Speakers on live streamEvery final (speakers: true)Streaming diarization +$0.0020/min PAYG
Live minutes / digestdigest: trueAudio Intelligence summarization (token add-on)
Languages60, auto-detectedNova-3 60+; Flux Multi 10 langs
Latency story~0.28 s first word; +~88 ms MTFlux marketed for ultra-low end-of-turn; Nova for ASR
Regions7 gateways; 2 s backlog capDeepgram cloud (+ EU endpoint noted on pricing)
Best cheap path on this cardNot the cheapest file STTPre-recorded Nova-3 Mono $0.0043/min

What 1,000 hours cost

WorkloadLecSyncDeepgram (PAYG list today)
Same-language live captions (mono)$360Flux EN / Nova-3 Mono streaming ~$462
Multilingual streaming STT$360 (any pair + MT)Flux Multi ~$468; Nova-3 Multi ~$552 (STT only)
Live captions + speakers$360Mono streaming + diarization ~$582
Translated live captions (arbitrary pair)$360STT + your own MT stack
Live minutesIncludedSeparate Audio Intelligence tokens
File / pre-recorded same-languageLive API not cheapest file pathNova-3 Mono pre-recorded ~$258

For flat-vs-add-on vendor math, see what 1,000 hours cost once translation, speakers and summaries are added. For stitching STT + MT yourself, see built-in translation vs a DIY pipeline.

When Deepgram wins

  • Voice agents that need Flux’s model-native turn detection and EagerEndOfTurn-style agent pipeline on Deepgram’s card.
  • Same-language file or batch ASR where pre-recorded Nova-3 Monolingual at $0.0043/min beats a live all-in caption API.
  • Growth prepaid volume where you already lock $4,000+/year credits and accept Growth streaming rates ($0.0065 Mono / Flux EN).
  • STT-only products where you already own translation, speakers and minutes as separate services.
  • Self-serve customization / keyterm workflows on Nova-3 with explicit add-on pricing.

When LecSync fits better

  • Live bilingual captions between any two of 60 languages on one socket and one minute price.
  • Speakers + audio-aligned timestamps on every final without a streaming diarization add-on line.
  • Running minutes (digest: true) while the session is still open.
  • Predictable all-in $/audio-min for captions + MT + speakers + digest at $0.006 ($360 per 1,000 hours).
  • Lecture and meeting capture where agent turn-taking is not the product and backlog must stay capped (2 s).

FAQ

How much does Deepgram streaming STT cost per minute today?

On deepgram.com/pricing (November 3, 2026), Pay As You Go streaming lists Flux English and Nova-3 Monolingual at $0.0077/min, Flux Multilingual at $0.0078/min, and Nova-3 Multilingual at $0.0092/min. Growth rates are lower ($0.0065–$0.0078). Promotional streaming prices from late October are not shown on today’s table.

How much does the LecSync Realtime Speech API cost?

LecSync bills $0.006 per accepted audio minute, all-in for realtime transcription, translation, speakers, timestamps, glossary, polish, live digest and optional 30-day storage. Turning features off does not lower the rate; turning them all on does not raise it. See developers pricing.

Did Deepgram’s public streaming prices change since the Flux comparison?

Yes for the promotional rows. The October 28, 2026 Flux post cited Flux English promo $0.0065/min and Nova-3 Monolingual streaming promo $0.0048/min. On November 3, 2026 the owned Streaming table shows the regular $0.0077 / $0.0092 PAYG figures without those promo labels.

Does Deepgram include live translation and speakers in the STT minute?

No for live any-to-any text translation on the STT streaming SKUs reviewed here. Streaming speaker diarization is listed as a $0.0020/min Pay As You Go add-on. Audio Intelligence summarization bills per token on top of transcription.

Is Deepgram cheaper than LecSync for file transcription?

Often yes for same-language pre-recorded Nova-3 Monolingual at $0.0043/min PAYG (Calculated ~$258 per 1,000 hours) versus LecSync’s live $0.006/min. Choose Deepgram when you only need file STT; choose LecSync when live MT, speakers and digest matter on one bill.

Which should I pick for live multilingual meeting captions?

Pick LecSync when you need bilingual captions, speakers and minutes on one WebSocket at $0.006/min. Pick Deepgram Nova-3 when same-language (or Nova multilingual STT without bundled MT) meeting transcription plus optional add-ons fits your stack, and budget the $0.0077–$0.0092/min streaming base plus diarization if you need speakers.

Bottom line

Deepgram’s November 3, 2026 public Streaming card posts $0.0077–$0.0092/min PAYG for Flux and Nova-3 — the late-October promo discounts are gone from the table. That STT card still wins for agents (Flux) and cheap pre-recorded Mono files ($0.0043/min). LecSync stays $0.006/min all-in when live captions need translation, speakers and digest without stacking add-ons. Start on the LecSync developers page, or compare the prior Flux and Nova write-ups for product detail beyond this rate refresh.