Deepgram Pricing Explained (2026): Speech-to-Text, Voice Agent API and What a Month Really Costs
Deepgram bills by the second of audio, with no minimums on Pay As You Go and $200 of free credit to start. Speech-to-text, text-to-speech and its all-in-one Voice Agent API are each priced separately. Every figure below is from deepgram.com/pricing (rate tables, plan comparison and FAQ), read October 21, 2026.
Deepgram's streaming prices are on promotion right now. Budget on the regular rate, add up the add-ons you'll switch on, and the bill won't surprise you when the promo ends.
Deepgram's Nova-3 streaming is $0.0048 a minute right now, at a limited-time promotional rate. Here's every Deepgram price line (speech-to-text, add-ons, text-to-speech and the Voice Agent API), what 1,000 to 100,000 minutes a month cost, and the free credit, Growth plan and concurrency rules that change the bill.
What you pay for on every Deepgram minute
Streaming speech-to-text
$0.0048–$0.0078/min
Pay As You Go: Nova-3 Monolingual $0.0048 (regular $0.0077), Nova-3 Multilingual $0.0058 (regular $0.0092), Flux English $0.0065 (regular $0.0077) and Flux Multilingual $0.0078. Deepgram marks the streaming rates as limited-time promotional prices. Flux is its model for voice agents, with built-in turn detection.
Add-ons
+$0.0013–$0.0020/min each
Only if you switch them on: redaction $0.0020, keyterm prompting $0.0013, entity detection $0.0017 and speaker diarization $0.0020 a minute on streaming (diarization is included on pre-recorded audio). Smart formatting is included.
Phone line
Not from Deepgram
Deepgram's pricing page lists no telephony. The call itself is billed by your carrier or voice platform.
Total per minute
about 0.5¢ to 0.8¢
Streaming transcription only, Pay As You Go, at today's promotional rates. Add-ons, text-to-speech and the Voice Agent API are separate.
What a month costs on Deepgram’s own numbers
1,000 min/month
- Nova-3 streaming ($0.0048)
- $4.80
- Flux English ($0.0065)
- $6.50
- Voice Agent Standard ($0.075)
- $75.00
- Voice Agent Advanced ($0.163)
- $163.00
5,000 min/month
- Nova-3 streaming ($0.0048)
- $24.00
- Flux English ($0.0065)
- $32.50
- Voice Agent Standard ($0.075)
- $375.00
- Voice Agent Advanced ($0.163)
- $815.00
10,000 min/month
- Nova-3 streaming ($0.0048)
- $48.00
- Flux English ($0.0065)
- $65.00
- Voice Agent Standard ($0.075)
- $750.00
- Voice Agent Advanced ($0.163)
- $1,630.00
25,000 min/month
- Nova-3 streaming ($0.0048)
- $120.00
- Flux English ($0.0065)
- $162.50
- Voice Agent Standard ($0.075)
- $1,875.00
- Voice Agent Advanced ($0.163)
- $4,075.00
50,000 min/month
- Nova-3 streaming ($0.0048)
- $240.00
- Flux English ($0.0065)
- $325.00
- Voice Agent Standard ($0.075)
- $3,750.00
- Voice Agent Advanced ($0.163)
- $8,150.00
100,000 min/month
- Nova-3 streaming ($0.0048)
- $480.00
- Flux English ($0.0065)
- $650.00
- Voice Agent Standard ($0.075)
- $7,500.00
- Voice Agent Advanced ($0.163)
- $16,300.00
Pay As You Go rates times minutes, with no platform fee. The first two columns are transcription only, at today's promotional rates (at the regular $0.0077, both cost $7.70 per 1,000 minutes). The Voice Agent API columns include speech-to-text, the LLM and text-to-speech, billed on websocket connection time. Add-ons and your phone line come on top, and new accounts spend the $200 free credit first. Want your own volume and carrier? Try the cost calculator.
Packages and add-ons
Pay As You Go
$200 free credit
No minimums, no expiration and no card required to start. Up to 150 streaming (WSS) and 50 REST speech-to-text connections, 5 for Deepgram Whisper Cloud, 45 for text-to-speech, 45 for the Voice Agent API and 10 for Audio Intelligence. Community and Discord support. Auto-load is on by default: it reloads $100 when your credit falls to $10, and you can turn it off.
Growth
$4K+/year
Pre-paid credits for the year, redeemed against usage, at up to 20% off. Up to 225 streaming connections and 60 for text-to-speech and the Voice Agent API. Overages are billed at your Growth rate plus 10%, every Monday, with a card on file; buying a new Growth plan early (“early renew”) avoids the premium.
Enterprise
Contact sales
Custom concurrency, volume discounts, custom model training and dedicated account management. Self-hosted containers for your own VPC or on-premise hardware (NVIDIA GPUs), and a BAA for customers handling health data.
Pre-recorded transcription
From $0.0043/min
Nova-3 Monolingual $0.0043 ($0.0036 on Growth), Nova-3 Multilingual $0.0052 ($0.0043), Whisper Large $0.0048. Speaker diarization is included. Older models stay available for existing deployments: Nova-2 streaming $0.35/hr, Enhanced $0.99/hr, Base $0.87/hr.
Text-to-speech
From $0.0150/1K characters
Billed per 1,000 characters of input text: Flux TTS $0.0450, Aura-2 $0.030, Aura-1 $0.0150 ($0.0405, $0.027 and $0.0135 on Growth). Until December 31, 2026, every $1 spent on Flux TTS earns $1 in credits, up to $500, on Pay As You Go and Growth.
Voice Agent API
$0.050–$0.163/min
Speech-to-text, LLM and text-to-speech in one, billed on websocket connection time. Pay As You Go: Standard $0.075, Standard with your own TTS $0.065, Custom with your own LLM $0.059, your own LLM and TTS $0.050, Advanced $0.163, Advanced with your own TTS $0.122. Growth: $0.068, $0.051, $0.041, $0.146 and $0.110 (no Growth rate listed for own LLM only).
Audio Intelligence
Per 1K tokens
Summarization $0.0003 per 1,000 input tokens and $0.0006 per 1,000 output tokens ($0.00024 and $0.00048 on Growth), on top of the transcription. Topic detection, sentiment analysis and intent recognition are listed alongside.
How it compares on published pricing
Deepgram- Starting price
- Nova-3 streaming $0.0048/min ($0.29/hr); $200 free credit
- Best for
- Streaming speech-to-text for voice agents
AssemblyAI- Starting price
- Streaming from $0.15/hr; $50 free credits
- Best for
- Low-cost streaming and batch transcription
Speechmatics- Starting price
- From $0.24/hr pay as you go (as low as ~$0.12/hr with discounts)
- Best for
- Multilingual transcription with language switching
Soniox- Starting price
- ~$0.10/hr async; ~$0.12/hr real-time
- Best for
- Very low-cost speech-to-text
Gladia
- Starting price
- Async from €0.16/hr; real-time €0.68/hr
- Best for
- Real-time multilingual transcription
Pricing explained, line by line:AssemblyAI pricing →
STT models compared
- Price
- Free until Nov 19, 2026; then $0.002/min
- Latency
- —
- Languages
- —
- Price
- $0.0065/min (English)
- Latency
- Built-in turn detection
- Languages
- English or multilingual
- Price
- $0.39 / hr
- Latency
- ~150 ms partials
- Languages
- 90+
- Price
- $0.0048/min streaming (monolingual, promo; reg. $0.0077)
- Latency
- —
- Languages
- 45+ (Nova models)
- Price
- $0.21 / hr
- Latency
- Async (pre-recorded)
- Languages
- 18
- Price
- $0.017/min
- Latency
- Live transcription
- Languages
- —
- Price
- $0.30 / hr real-time
- Latency
- Built-in turn detection and VAD (real-time)
- Languages
- 55+
- Price
- $0.40 / hr real-time; $0.24 / hr files
- Latency
- Real-time or files
- Languages
- 56, switches mid-sentence
- Price
- $0.30 / hr files
- Latency
- Files (real-time coming soon)
- Languages
- 55+, tuned for clinical terms
- Price
- ~$0.12 / hr real-time; ~$0.10 / hr async
- Latency
- Real-time streaming
- Languages
- 60+ (translation included)
- Price
- €0.16 / hr async (real-time €0.68 / hr)
- Latency
- Async (files)
- Languages
- 100+
See all voice models · Latency and prices as each vendor publishes them.
What to check before you choose
- Budget on the regular price. Deepgram labels its streaming rates limited-time promotional prices: Nova-3 Monolingual is $0.0048 now against $0.0077 regular, Multilingual $0.0058 against $0.0092, and Flux English $0.0065 against $0.0077. At 10,000 minutes a month, Nova-3 Monolingual is $48 today and $77 at the regular rate.
- Nova-3 or Flux? Flux is built for voice agents, with turn detection built in, and costs $0.0017 a minute more than Nova-3 at today's prices ($17 per 10,000 minutes). At regular prices the English versions are the same $0.0077.
- Add-ons can double the bill. On streaming, redaction and speaker diarization add $0.0020 a minute each, so turning on both adds $0.0040 to a $0.0048 Nova-3 minute. Diarization is included on pre-recorded audio.
- The Voice Agent API is the bundled option. Standard at $0.075 a minute covers listening, the LLM and speaking, and bringing your own LLM or voice lowers the rate (down to $0.050 with both). It's billed on websocket connection time, and the phone line isn't included. ElevenLabs Agents is $0.08 a minute plus the LLM, and Vapi's calculator puts its default stack at about 8.2 to 12.9¢ before carrier fees.
- Check concurrency before launch. Pay As You Go allows up to 150 streaming connections and 45 for the Voice Agent API; Growth lifts that to 225 and 60. Deepgram's FAQ says requests over your limit may be queued or rejected.
- Compare per hour against other speech APIs. Nova-3 streaming is $0.29 an hour today. On their own pricing pages, re-checked today, AssemblyAI streaming starts at $0.15 an hour, Soniox real-time is about $0.12, Speechmatics starts at $0.24 an hour (files) and $0.30 real-time, and Gladia real-time is €0.68 an hour. For trying them out, Deepgram's $200 free credit compares with $50 at AssemblyAI, $100 at Speechmatics and €50 at Gladia.
Questions people ask
How much does Deepgram cost per minute?
Speech-to-text starts at $0.0048 a minute ($0.29 an hour) for Nova-3 streaming on Pay As You Go, a limited-time promotional rate (regular $0.0077). Flux, Deepgram's voice agent model with built-in turn detection, is $0.0065 a minute in English. The Voice Agent API, which bundles speech-to-text, the LLM and text-to-speech, starts at $0.075 a minute on its Standard tier. Billing is per second, and new accounts get $200 in free credit.
Is Deepgram free to start?
Yes. Every new account gets $200 in free credit with no card required, and it doesn't expire. Deepgram says that's about 43,000 minutes of Nova transcription. At the Voice Agent API's Standard rate of $0.075 a minute, it covers about 2,667 minutes. Promotional and free credits can't be transferred between accounts.
How much does Deepgram's Voice Agent API cost?
On Pay As You Go: Standard $0.075 a minute, Standard with your own text-to-speech $0.065, Custom with your own LLM $0.059, your own LLM and text-to-speech $0.050, Advanced $0.163 and Advanced with your own text-to-speech $0.122. It's billed on websocket connection time, and Growth rates are lower (Standard $0.068).
What's the difference between Pay As You Go and Growth?
Pay As You Go has no commitment and bills on usage. Growth starts at $4,000 a year in pre-paid credits and gives up to 20% off and more concurrency (225 streaming connections against 150). Going over your Growth credits costs your Growth rate plus 10%, billed weekly, unless you renew early.
How is Deepgram text-to-speech priced?
Per 1,000 characters of input text: Aura-1 $0.0150, Aura-2 $0.030 and Flux TTS $0.0450 on Pay As You Go, slightly less on Growth. Until December 31, 2026, Deepgram adds $1 in credits for every $1 spent on Flux TTS, up to $500.
Does Deepgram round up or charge for silence?
Deepgram says it uses true per-second billing, so a 14-second file is billed as 14 seconds. Multichannel audio is billed per channel: a 10-minute stereo file counts as 20 minutes.
Is Deepgram HIPAA compliant?
Deepgram says it is SOC 2 Type 1 and Type 2 certified and HIPAA compliant, and it signs Business Associate Agreements for Enterprise customers. It also offers an EU endpoint (api.eu.deepgram.com) and self-hosted deployment on Enterprise.
Ready to look closer? See Deepgram’s full listing and price history → Or weigh the options in Deepgram Alternatives for Speech-to-Text Compared.
Not sure which fits? Answer a few quick questions in the finder →
Prices from vendors' official pricing pages, checked October 21, 2026. Not sponsored; order isn't a ranking.