A buyer who searches for Deepgram Aura is reading about a model Deepgram now calls previous-generation. Deepgram's newer text to speech model is Flux TTS, which Deepgram recommends for all new builds, and the two bill differently: Aura-2 costs $30 per million characters on pay as you go, and Flux TTS costs $45, which is more than Bland Speech's $40.
Deepgram also sells speech to text, and its text to speech sits beside that product for teams that want both from one vendor. That is a sound reason to buy it, though a team whose callers will judge the voice has more to weigh than convenience.
What is Deepgram Aura, and how does it work?
Deepgram Aura is Deepgram's text to speech line, which now runs across three generations: Aura, Aura-2, and Flux TTS. Deepgram describes Flux TTS as its latest-generation model, built to be conversation-native, streaming-first, and expressive by default, and as of November 2026 it calls Aura-2 its previous-generation model with voices across seven languages. The first-generation Aura has English voices only. Aura and Aura-2 are served on /v1/speak, and Flux TTS is served on /v2/speak.
Flux TTS runs two ways: a WebSocket on /v2/speak for real-time conversation, and a batch REST call to the same path. Deepgram says Flux TTS responses start in as low as 80ms at full 24kHz fidelity, even under production load, and Bland publishes 199ms to first audio. Each figure is the company's own claim.
What features does Deepgram Aura add?
Flux TTS has a delivery dial from -2 for calm to 2 for animated, marked beta, and a speed control from 0.5 to 1.5 in steps of 0.05 that changes without reconnecting. Deepgram says conversational state persists across turns, so tone carries forward from one reply to the next in a call.
Markup is narrower than a team coming from Amazon or Microsoft might expect. Flux TTS strips SSML, the XML markup older engines use for pauses and emphasis, and returns an INPUT_MARKUP_STRIPPED warning while synthesis continues. Inline pronunciation by IPA, the phonetic alphabet, is in Early Access on both transports, and an inline pause works on batch requests only.
On languages, Deepgram says Flux TTS ships in English plus 10 additional languages with built-in text normalization, and supports 40+ more for teams that provide their own normalization layer. Phone output comes as mu-law or A-law at 8,000 Hz by default, or 16,000 Hz, and Deepgram's Twilio guide says requesting mu-law at 8 kHz matches Twilio's Media Streams format exactly.
How easy is Deepgram Aura to use?
Deepgram Aura is easy to start with and gets more involved once a team is choosing between generations. A developer who already uses Deepgram for transcription can stay with one vendor, and Deepgram publishes guides for Pipecat, LiveKit Agents, and Twilio. Pipecat has a Deepgram text to speech service, and LiveKit offers Deepgram plugins for Python and Node.js.
The catch is the version choice. Aura-2 is cheaper, Flux TTS is the one Deepgram recommends, and the two use different endpoints, so moving later means changing code as well as a model name.
How much does Deepgram Aura cost?
Deepgram Aura costs $0.030 per 1,000 characters for Aura-2 on pay as you go, $0.0450 for Flux TTS, and $0.0150 for first-generation Aura, all billed per 1,000 characters of input text. Growth, a plan of prepaid credits from $4,000 a year that Deepgram says saves up to 20%, lowers Aura-2 to $0.027 and Aura-1 to $0.0135, and leaves Flux TTS at $0.0450. Enterprise is custom.
A free account gets $200 of credit with no card and no expiry. Through 31 December 2026, Deepgram adds $1 in credits for every $1 spent on Flux TTS, up to $500. Past a Growth plan's credits, Deepgram bills at the Growth rate plus 10%, charged weekly on Mondays.
What does a month of speech cost on Bland Speech and on Deepgram?
Bland Speech costs $0.04 per 1,000 characters, which is $40 per million, with no monthly plan, and a new account gets 50,000 characters free once. Enterprise pricing is custom, set with Bland's sales team.
Aura-2 is cheaper than Bland Speech at every volume, and Flux TTS costs more at every volume, which puts the question back on what each voice sounds like.
What these rows mean
- Bland Speech: $0.04 per 1,000 characters, one standard rate in Bland's studio or its API, with no plan fee.
- Deepgram Flux TTS: $0.0450 per 1,000 characters on pay as you go, the same on Growth.
- Deepgram Aura-2: $0.030 per 1,000 characters on pay as you go. Growth's $0.027 needs a prepaid plan from $4,000 a year, so it is left out. Free credit, promotions, taxes, and phone usage are left out of every column.
Is Deepgram Aura secure enough for enterprise calls?
Deepgram says it holds SOC 2 Type 1 and Type 2 certification, and it describes itself as HIPAA compliant, PCI compliant, and fully compliant with GDPR, with a dedicated EU endpoint for data processing inside the European Union. Deepgram signs BAAs for Enterprise customers handling electronic protected health information, and says BAAs are available for eligible customers of its text to speech.
Deployment is a real strength. Deepgram says Flux TTS supports self-hosted deployment, run in a customer's own cloud, private cloud, or on-premise infrastructure, for teams with data residency, security, or compliance requirements. A team that must keep audio inside its own network has that route with Deepgram.
Bland is SOC 2 Type II audited, HIPAA compliant with a signed BAA, and FedRAMP 20x Class C certified at Moderate, granted on 17 November 2026, and its trust page says speech synthesis runs inside the FedRAMP boundary. Bland runs its own model and speech, with zero AI sub-processors in the call path. Enterprise customers get dedicated orchestration and dedicated GPUs run by Bland.
Which sounds more human, Deepgram Aura or Bland Speech?
Deepgram Aura wasn't in Design Arena's Audio Realism Bench results of 4 August 2026, so there is no score from that test to set against Bland Speech's. In those results Bland Speech v3 ranked first of 13 AI voices at 1365, and a human recording scored 1500.
- #1BlandBland Speech v31365
- 2ndMicrosoftMAI-Voice-21224
- 3rdxAIGrok TTS1156
- 4thOpenAIGPT Realtime 21119
- 5thGoogleGemini 2.5 Pro TTS1109
- Not rankedDeepgram's own voicesNot ranked
Ranks are among 13 AI voices, and Deepgram's own voices weren't ranked. Real human recordings scored 1500, at the right edge of this scale. Bars start at 600, not zero. Source: Design Arena's published results.
Listeners in that test hear two clips of the same line from two models, without being told which is which, and pick the one that sounds more like a real person. Deepgram calls Flux TTS expressive by default, which is a claim a listening test can check, and a team that wants to check it can play both voices on the same script.
Is Deepgram Aura worth it, or should a team choose Bland Speech?
Deepgram Aura is worth it for a team that already buys Deepgram transcription, wants to run the model on its own infrastructure, or wants a low per-character rate, and Bland Speech is the better choice for a phone line where callers judge the voice. A team weighing more vendors can read Deepgram Aura alternatives.
Deepgram Aura pros and cons
Pros
- First audio in as low as 80ms on Flux TTS, as Deepgram states it
- Aura-2 at $30 per million characters and first-generation Aura at $15
- Speech to text and text to speech from one vendor, with Pipecat, LiveKit, and Twilio guides
- Deployment in a customer's own cloud, private cloud, or on-premise infrastructure
- Mu-law and A-law at 8,000 Hz for phone lines
- $200 of free credit with no card and no expiry
Cons
- Aura-2 is previous-generation, and Deepgram recommends Flux TTS for all new builds
- Flux TTS costs $0.0450 per 1,000 characters, above Bland Speech's $0.04
- Not in Design Arena's Audio Realism Bench results of 4 August 2026, so no score from that test
- Flux TTS strips SSML, and its delivery dial and inline pronunciation are beta or Early Access
Why do enterprises choose Bland Speech over Deepgram Aura?
Enterprises choose Bland Speech when the voice is what a caller judges and when a team would rather not pick between model generations. Bland Speech v3 ranked first on Design Arena's listening test of 4 August 2026, and there is no version choice to make: the voice a team passes decides the model, and btts-3 is the current one.
A voice agent streams a language model's words as they arrive, and Bland Speech takes them over WSS /v2/tts/ws, where a new context interrupts the turn that is playing when a caller cuts in. Output is mu-law at 8 kHz or PCM from 8 to 48 kHz. Bland Speech has a first-party Pipecat service and a LiveKit Agents plugin for Python, and an OpenAI-compatible endpoint, POST /v2/audio/speech, lets a team switch by changing a base URL, a key, and a model name. The text to speech API page has the code.
Performance tags such as [laughs], [sighs], or [say warmly] tell the model how a line should sound. An instant clone needs one clean sample of about 10 seconds, and a professional clone is fine-tuned on 30 minutes or more of audio after Bland's consent and rights review, with every clone private to the customer's organization. The Bland Speech page has a studio to try them.
Pros
- Bland Speech v3 ranked first of 13 AI voices on Audio Realism Bench on 4 August 2026
- One standard rate, $0.04 per 1,000 characters, in the studio or the API
- An OpenAI-compatible endpoint, so an existing integration changes a base URL, a key, and a model name
- Performance tags for how a line is read
- Mu-law at 8 kHz and PCM from 8 to 48 kHz for phone lines
- Instant clones from about 10 seconds of audio and professional clones from 30 minutes or more
Cons
- More than a hobby project needs, because Bland Speech is built for production calls
Verdict: Deepgram Aura or Bland Speech for enterprise calls?
For enterprise phone calls where the voice should sound like a person, Bland Speech is the pick, with Design Arena's 4 August 2026 first place for Bland Speech v3 and one standard rate behind it. Deepgram is the pick for a team that wants to keep speech to text and text to speech with one vendor, run the model on its own infrastructure, or pay a lower rate per character with Aura-2. More detail sits in the Bland Speech and Deepgram comparison, and the Cartesia vs Deepgram Aura page puts two rivals side by side. The text to speech reviews collection has every vendor's review, and the text to speech comparison hub links the rest.
Frequently asked questions
What is the difference between Deepgram Aura and Flux TTS?
Aura-2 is Deepgram's previous-generation model, served on /v1/speak with voices across seven languages, and Flux TTS is its latest-generation model, served on /v2/speak and recommended for all new builds. Flux TTS costs $0.0450 per 1,000 characters against $0.030 for Aura-2 on pay as you go.
How much does Deepgram Aura cost?
Aura-2 costs $0.030 per 1,000 characters on pay as you go and $0.027 on Growth, Flux TTS costs $0.0450 on both, and first-generation Aura costs $0.0150 and $0.0135. Bland Speech costs $0.04 per 1,000 characters with no plan.
Does Deepgram Aura support SSML?
Flux TTS strips recognized SSML tags and returns an INPUT_MARKUP_STRIPPED warning while synthesis continues. Flux TTS takes inline pronunciation by IPA in Early Access, and an inline pause on batch requests only. Bland Speech takes bracketed performance tags such as [laughs] and timed pauses written as <|0.5|>.
Is Deepgram Aura HIPAA compliant?
Deepgram says it is HIPAA compliant and signs BAAs for Enterprise customers handling electronic protected health information. Bland is HIPAA compliant with a signed BAA.
Can Deepgram Aura be self-hosted?
Yes. Deepgram says Flux TTS supports self-hosted deployment in a customer's own cloud, private cloud, or on-premise infrastructure. Bland's answer for dedicated capacity is that Enterprise customers get dedicated orchestration and dedicated GPUs run by Bland.
Is Deepgram Aura faster than Bland Speech?
By each company's published figure, yes. Deepgram says Flux TTS responses start in as low as 80ms, and Bland publishes 199ms to first audio. Neither figure comes from a test a third party ran.
Does Deepgram Aura work with Pipecat and LiveKit?
Yes. Pipecat has a Deepgram text to speech service, and LiveKit offers Deepgram plugins for Python and Node.js. Bland Speech has a first-party Pipecat service and a LiveKit Agents plugin for Python.
What is the best Deepgram Aura alternative?
Bland Speech is the best Deepgram Aura alternative for phone calls and voice agents, because it ranked first of 13 AI voices on Audio Realism Bench on 4 August 2026 and bills $0.04 per 1,000 characters. The text to speech API pricing comparison puts every rate side by side.
Sources
Deepgram's pricing page, text to speech product page, developer documentation on models, languages, Flux TTS features, and output formats, and its changelog. Pipecat's and LiveKit's text to speech service lists. Bland's pricing and trust and security pages and Bland Speech documentation. Design Arena's Audio Realism Bench article of 4 August 2026.
