Deepgram $200 Free Credit: No Card Required (2026)
$200 free credit for new accounts, no credit card
Every new Deepgram account gets $200 in free credit on the Pay As You Go plan, with no credit card required. Deepgram says this is about 43,000 minutes (over 700 hours) of Nova transcription, and the credit works across speech-to-text, text-to-speech, the Voice Agent API and audio intelligence. Deepgram's pricing page says the credit lasts until used, but its developer docs say free promotional credits expire one year from signup.
Who is eligible
- New Deepgram account
- No credit card needed to sign up
How to claim it
- Sign up at console.deepgram.com
- The $200 credit is attached to your first project automatically
- Create an API key in that project and call speech-to-text, text-to-speech or Voice Agent APIs
Limits
| Credit | $200 on Pay As You Go |
|---|---|
| Deepgram's estimate | ~43,000 minutes (700+ hours) of Nova transcription |
| Speech-to-text concurrency | 50 REST / 150 streaming (WSS) |
| Whisper Cloud concurrency | 5 |
| Text-to-speech concurrency | 45 |
| Voice Agent concurrency | 45 |
| Audio intelligence concurrency | 10 |
Catches & caveats
Deepgram's pricing page and developer docs disagree on expiry: the pricing page says the $200 is available until used, while the docs say free promotional credits expire one year from signup. We show both and suggest checking the expiry in your console.
Deepgram gives every new account $200 in free credit on its Pay As You Go plan, and you don't need a credit card to get it. Deepgram says that is roughly 43,000 minutes (over 700 hours) of transcription with its Nova model. The credit is not limited to speech-to-text: it also covers text-to-speech, the Voice Agent API and audio intelligence. The one point to watch is expiry. The pricing page says the credit lasts until you use it, but Deepgram's developer documentation says free promotional credits expire one year from signup.
What the $200 covers
The Pay As You Go plan includes all endpoints in Deepgram's public models, and the free credit is spent at the same Pay As You Go rates as paid usage. These are the main rates on Deepgram's pricing page in October 2026:
| Product | Pay As You Go rate | What $200 buys (approx.) |
|---|---|---|
| Nova-3 monolingual, pre-recorded | $0.0043/min ($0.26/hr) | ~769 hours |
| Nova-3 multilingual, pre-recorded | $0.0052/min ($0.31/hr) | ~645 hours |
| Nova-3 monolingual, streaming | $0.0048/min ($0.29/hr) | ~689 hours |
| Flux English, streaming | $0.0065/min ($0.39/hr) | ~512 hours |
| Flux multilingual, streaming | $0.0078/min ($0.47/hr) | ~425 hours |
| Aura-2 text-to-speech | $0.030 per 1,000 characters | ~6.7M characters |
| Aura-1 text-to-speech | $0.015 per 1,000 characters | ~13.3M characters |
| Flux TTS | $0.045 per 1,000 characters | ~4.4M characters |
| Voice Agent API, Standard | $0.075/min ($4.50/hr) | ~44 hours of agent time |
| Voice Agent API, Advanced | $0.163/min ($9.75/hr) | ~20 hours of agent time |
The right-hand column is our arithmetic: $200 divided by the hourly or per-character rate. For example, $200 ÷ $0.26 per hour = 769 hours of Nova-3 pre-recorded audio, and $200 ÷ $0.030 per 1,000 characters = 6.67 million characters of Aura-2 speech. Add-ons reduce these numbers.
Add-ons and audio intelligence
Speech-to-text features are priced on top of the base rate. On pre-recorded audio, Deepgram lists redaction at $0.0020/min and keyterm prompting at $0.0013/min on Pay As You Go, while speaker diarization is included; other add-ons such as entity detection have their own per-minute rates in the same table. Audio intelligence features (summarization, topic detection, sentiment analysis and intent recognition) are billed per token: $0.0003 per 1,000 input tokens and $0.0006 per 1,000 output tokens on Pay As You Go.
Billing details that stretch the credit
- Per-second billing. Deepgram's FAQ says a 14-second file is billed as 14 seconds, not rounded up to 15 seconds or a minute.
- Multichannel counts per channel. A 10-minute stereo file processed as two channels is billed as 20 minutes.
- Text-to-speech is billed per 1,000 characters of input text for Flux TTS, Aura-2 and Aura-1.
- Voice Agent API is billed per minute of WebSocket connection time, so idle connections still cost money.
Does the credit expire?
This is the one area where Deepgram's own pages disagree, so we show both:
- Pricing page: the Pay As You Go plan is described as "No minimums. No expiration." The FAQ says that unlike free tiers that expire after 12 months, the $200 credit is available until you use it up.
- Developer documentation: the Managing Projects guide says Deepgram free promotional credits expire one year from signup, while credits bought by individuals with a credit card do not expire, and enterprise-contract credits expire at the end of the contract.
Until Deepgram aligns these pages, plan as if the $200 expires 12 months after signup. Your console's billing view shows the balance on your project; if an expiry date is shown there, that is the one that applies to your account.
Who qualifies
Deepgram's pricing page says every new account receives the $200 credit. It does not publish any other eligibility rules, such as country restrictions, business verification or student status. No credit card is required. The credit is a sign-up offer, so creating extra accounts to collect it again is not something Deepgram describes as allowed.
How to claim it, step by step
- Go to console.deepgram.com/signup and create an account.
- Deepgram structures accounts into projects. Your first project is created for you, and any promotional credit is attached to it.
- In that project, create an API key. Deepgram recommends separate keys for testing and production.
- Make a request to the speech-to-text, text-to-speech or Voice Agent API. Usage is deducted from the $200 balance automatically.
- Check usage and remaining balance in the console as you test.
curl -X POST "https://api.deepgram.com/v1/listen?model=nova-3&smart_format=true" \
-H "Authorization: Token $DEEPGRAM_API_KEY" \
-H "Content-Type: audio/wav" \
--data-binary @meeting.wavCredit lives in the first project. If you create a second project for production, its usage won't draw on the $200 unless you ask Deepgram Support to transfer the balance. Deepgram's docs say any expiry date moves with transferred credit.
Limits on the free credit
The free credit uses Pay As You Go limits. These are the concurrency limits on the pricing page; Deepgram says requests over the limit may be queued or rejected.
| Product | Pay As You Go | Growth |
|---|---|---|
| Speech-to-text, REST | 50 | 50 |
| Speech-to-text, streaming (WSS) | 150 | 225 |
| Speech-to-text, Whisper Cloud | 5 | 5 |
| Text-to-speech, REST + WSS | 45 | 60 |
| Voice Agent, WSS | 45 | 60 |
| Audio intelligence, REST | 10 | 10 |
Deepgram's docs also note that API keys work only inside the project they were created in, the account that created a key must stay in that project for the key to keep working, and programmatically generated API keys are limited to 250 per day.
Older models and the credit
Deepgram still serves its older speech-to-text models for existing deployments at unchanged rates: Nova-2 streaming at $0.35 per hour, Enhanced at $0.99 per hour and Base at $0.87 per hour, with Growth rates about 12.5% lower. The free credit is spent at whatever rate the model you call charges, so testing on Enhanced or Base uses the $200 roughly three to four times faster than Nova-3 pre-recorded at $0.26 per hour. Deepgram recommends that new projects start with its current models.
A sensible test plan for the $200
Because the credit spans several products, it is worth deciding up front what you want to learn. One way to split it for a voice product prototype:
- Accuracy test (about $10): run 30 to 40 hours of your own recordings through Nova-3 pre-recorded and compare transcripts with a human reference.
- Latency test (about $20): stream 50 or more hours through Nova-3 or Flux streaming to measure real-time behaviour on your network.
- Voice output (about $20): generate around 650,000 characters with Aura-2 to compare voices.
- Agent pilot (the rest): run end-to-end Voice Agent calls, which are the most expensive per minute at $0.075 on the Standard tier.
These amounts are illustrative arithmetic at the published rates, not Deepgram recommendations.
What happens after the $200
When the credit is used up you stay on Pay As You Go: no minimums, billed monthly on usage. To keep going you add a payment method in the console. For larger volumes, Deepgram offers:
- Growth: prepaid annual credits starting at $4,000 a year, with rates up to about 20% lower (for example Nova-3 pre-recorded at $0.0036/min instead of $0.0043), higher streaming concurrency and priority support. Overages are billed at the Growth rate plus a 10% premium, and a card must be on file.
- Enterprise: custom pricing, deployment options including on-premise or private cloud, and custom limits.
Worked cost examples after the credit
- A podcast app transcribing 300 hours a month of English audio with Nova-3 pre-recorded: 300 × $0.26 = $78 a month. The $200 credit covers the first two and a half months.
- A voice agent handling 2,000 five-minute calls a month on the Standard Voice Agent API: 10,000 minutes × $0.075 = $750 a month. The credit covers about 2,667 minutes, roughly the first eight days.
- Reading 1 million characters of text aloud a month with Aura-2: 1,000 × $0.030 = $30 a month, so the credit lasts more than six months.
Getting the most out of the credit
- Pick the model for the job. Nova-3 monolingual is the cheapest current speech-to-text option; use multilingual or Flux only where you need them.
- Use pre-recorded for batch work. It is cheaper per minute than streaming ($0.0043 vs $0.0048 for Nova-3 monolingual).
- Merge channels when you don't need them separate. Multichannel processing bills each channel.
- Close idle Voice Agent connections. Billing is on WebSocket time.
- Turn on only the add-ons you use. Redaction and keyterm prompting, for example, each add to the per-minute cost.
How it compares with other free speech APIs
| Provider | Free offer | Notes |
|---|---|---|
| Deepgram | $200 credit, no card | Covers STT, TTS and Voice Agent |
| AssemblyAI | $50 credit, no card | Free accounts limited to 5 concurrent pre-recorded jobs |
| ElevenLabs free plan | 10,000 credits a month | Mainly text-to-speech; non-commercial with attribution |
| Groq free tier | Rate-limited free plan | Includes Whisper speech-to-text |
| Cloudflare Workers AI | 10,000 Neurons a day | Runs speech models inside Workers |
For startups, Deepgram for Startups is a separate program offering up to $100,000 in credits to be used within 12 months. Deepgram says applicants should be early-stage startups in production or launching within six months, and that responses typically arrive within a week. The ElevenLabs equivalent is its startup grants program, and our guide to getting free AI credits covers other startup programs.
Common mistakes
- Assuming the credit never expires. Deepgram's documentation says promotional credits expire a year after signup, even though the pricing page says otherwise.
- Creating a new project and wondering why the balance isn't used. Credit is attached to the first project.
- Leaving streaming or agent sockets open. Streaming and Voice Agent usage is billed on connection time.
- Forgetting that stereo doubles the bill. Each processed channel is billed separately.
Frequently asked questions
›Does Deepgram give free credits?
Yes. Every new account gets $200 in free credit on the Pay As You Go plan, which Deepgram says is about 43,000 minutes of Nova transcription.
›Does the Deepgram $200 credit expire?
Deepgram's sources disagree. The pricing page says the credit is available until you use it up, while the developer docs say free promotional credits expire one year from signup. Plan to use it within 12 months and check the expiry shown in your console.
›Do I need a credit card for Deepgram?
No. The pricing page says Pay As You Go requires no credit card.
›What can I use the $200 credit on?
The Pay As You Go plan includes all endpoints in public models, so the credit covers speech-to-text (Nova-3, Flux), text-to-speech (Aura-2, Aura-1, Flux TTS), the Voice Agent API and audio intelligence features.
›How many hours of transcription does $200 buy?
Deepgram says about 43,000 minutes (over 700 hours) using its Nova model. At today's Nova-3 pre-recorded rate of $0.26 per hour, $200 ÷ $0.26 is about 769 hours; streaming at $0.29 per hour is about 689 hours.
›Can I move the free credit to another project?
Promotional credit is attached to your first project. Deepgram's docs say you can contact Support to transfer it, and any expiry date moves with it.
›What happens when the $200 runs out?
You continue on Pay As You Go, billed monthly for usage with no minimums, or you can move to the Growth plan, which starts at $4,000 a year in prepaid credits for up to about 20% lower rates.
›Is there a bigger free credit for startups?
Yes. Deepgram for Startups offers up to $100,000 in API credits that must be used within 12 months, for early-stage startups in production or launching within six months.
Sources
Checked against these pages on Oct 17, 2026. Providers change offers often: always confirm on the official page before you rely on it.
- Official: Deepgram pricing(deepgram.com)
- Official: Deepgram docs, Managing projects (credits and expiry)(developers.deepgram.com)
- Official: Deepgram startup program(deepgram.com)
- Official: Deepgram console signup(console.deepgram.com)
Similar deals
Google Cloud Free Trial
$300 Welcome credit valid for 90 days
New Google Cloud customers get a $300 Welcome credit to spend over a 90-day Free Trial on most Google Cloud services. You must add a valid payment method, but you are not billed during the trial. The credit cannot pay for the Gemini API in AI Studio or for partner models offered as managed APIs, and trial accounts cannot add GPUs, use Marketplace or request quota increases. If you don't upgrade, resources are stopped at the end and deleted after a 30-day grace period. Separately, the always-free Free Tier keeps running with no end date.
AssemblyAI Speech-to-Text API
$50 free credits, no credit card
AssemblyAI gives new accounts $50 in free credits with no credit card required. AssemblyAI's pricing FAQ says this covers up to 185 hours of pre-recorded or 333 hours of streaming transcription, depending on the model. Free accounts can use pre-recorded transcription, Speech Understanding and streaming, but not the LLM Gateway, and have lower concurrency limits until they upgrade.
Gemini API / Google AI Studio
Free tier: Gemini 3.8 Flash, 3.6 Flash, 3.5 Flash-Lite, 2.5 Pro, 2.5 Flash, Gemma 4 and more, rate-limited
Google's Gemini API has a Free tier where many models are listed as 'Free of charge' for input and output, including Gemini 3.8 Flash, Gemini 3.6 Flash, Gemini 3.5 Flash-Lite, Gemini 2.5 Pro, Gemini 2.5 Flash, Gemini Embedding 2 and Gemma 4. Newer flagship, image, video and music models (Gemini 4 Argon, Gemini 3.1 Pro Preview, Nano Banana, Veo 3.1, Lyria) are paid-only. Free-tier prompts and responses may be used to improve Google's products and can be read by human reviewers. Rate limits apply per project and are shown in Google AI Studio rather than published as a fixed table.
GroqCloud API
Free plan: rate-limited access to models like gpt-oss-120b, Qwen3.8-27B and Whisper
GroqCloud has a Free plan with per-model rate limits published on its rate-limits docs page, for models including gpt-oss-120b, gpt-oss-20b, Qwen3.8-27B, Whisper and Orpheus text-to-speech. The paid Developer plan is pay-as-you-go with much higher limits (for example 1K RPM instead of 30 for gpt-oss-120b) plus Batch and Flex processing. Limits apply at the organization level and your exact numbers are shown on the console Limits page.