Last verified October 10, 2026.
We make ReadAloud, so it appears below as one of the options, and we say so. Options are listed alphabetically, not ranked.
Who Deepgram Aura suits
We file Deepgram Aura under text-to-speech API platforms: companies whose main product is a speech API that developers call from their own code. It lists a price per character. The audience its pages name: teams already using Deepgram for transcription or its Voice Agent API who want TTS on the same platform, plus developers who price on the Aura tier ($15 per 1M characters) or want multiple European languages and Japanese (Aura-2).
How Deepgram Aura prices
In Deepgram Aura's words: Pay-as-you-go per character, with discounted Growth plan rates for annual prepaid commitments.
- Flux TTS (Pay As You Go): $45 per 1M characters. English-only today, served on /v2/speak; $45 also confirmed on the Flux product page. (source: deepgram.com)
- Aura-2 (Pay As You Go): $30 per 1M characters. Seven languages; $30 also confirmed on the Aura-2 launch article. Aura-2 is the widest-language model; Deepgram recommends Flux TTS (English-only) for new builds, listed at $45. (source: deepgram.com)
- Aura (Pay As You Go): $15 per 1M characters. First-generation, English-only voices. Single-source (pricing page). (source: deepgram.com)
- Flux TTS (Growth): $40.50 per 1M characters. Growth plan: $4K+/year prepaid credits. Single-source. (source: deepgram.com)
- Aura-2 (Growth): $27 per 1M characters. Growth plan. Single-source. (source: deepgram.com)
- Aura (Growth): $13.50 per 1M characters. Growth plan. Single-source. (source: deepgram.com)
Check whether the listed rate is for the model and latency class you would actually call, and whether concurrency or support tiers cost extra.
Every per-character list price we found for Deepgram Aura, Flux TTS (Pay As You Go) at $45, Aura-2 (Pay As You Go) at $30, Aura (Pay As You Go) at $15, Flux TTS (Growth) at $40.50, Aura-2 (Growth) at $27 and Aura (Growth) at $13.50, is higher than both ReadAloud Live ($4) and ReadAloud Studio ($10) per 1M characters.
Free tier at Deepgram Aura: $200 free credit on sign-up, no credit card required, then pay-as-you-go (credit applies to all products, not specific to TTS).
What Deepgram Aura lists as strengths
- Aura-2 supports seven languages; Flux TTS is a newer English model built for streaming voice agents with interruption reporting. (source: developers.deepgram.com)
- Vendor states Flux TTS output starts in as low as 80 ms (vendor-stated). (source: deepgram.com)
- Offers speech-to-text, TTS and a Voice Agent API on one platform with a single account and API key. (source: deepgram.com)
- Publishes SOC 2, HIPAA and GDPR statements and a dedicated EU endpoint. (source: deepgram.com)
- Entry prices: Aura (first generation) is $15 per 1M characters and Aura-2 is $30 per 1M, with a $200 free credit. (source: deepgram.com)
Conditions to check with Deepgram Aura
- Flux TTS (the recommended newer model) is English-only today; other languages require Aura-2. (source: developers.deepgram.com)
Streaming and time to first audio
Deepgram Aura: WebSocket streaming is listed and HTTP streaming is listed. Vendor-stated: Flux TTS output starts 'in as low as 80 ms'. Not independently measured. Test from your own servers with your own text.
Voices and languages
Deepgram Aura: Roughly 36 Flux TTS English voices, about 90 Aura-2 voices across 7 languages and 12 first-generation Aura English voices (counted by us from the docs voice catalog page text, so approximate); languages listed: Flux TTS: English only (several accents). Aura-2: English, Spanish, German, French, Dutch, Italian, Japanese. Aura: English.
Limits and formats
- Concurrency: Pay As You Go, North America: Aura and Aura-2 up to 15 concurrent REST and 45 concurrent streaming requests; Flux TTS 15 REST / 45 streaming in North America and 5 in other listed regions. (source: developers.deepgram.com)
- Output formats listed: linear16, mulaw, alaw, mp3 (REST), opus (REST), flac (REST) and aac (REST).
API shape and switching cost
We could not tell whether Deepgram Aura offers an OpenAI-compatible speech endpoint. Deepgram uses its own /v1/speak (Aura) and /v2/speak (Flux TTS) endpoints with 'Authorization: Token' auth and the voice chosen via the model query parameter; no OpenAI-style speech route was found in the pages read.
What we could not confirm about Deepgram Aura
- Maximum characters per request
- Custom voice or cloning availability
- Commercial-use terms for TTS output (not stated on pages read)
- SSML support (not documented in pages read; Flux page says no SSML needed)
- Aura (first-gen) and Growth-plan prices are single-source (pricing page)
- Whether an OpenAI-compatible speech route exists
- Voice counts are approximate, counted from the docs catalog text
- Time-limited offers (a credit match and a free build period for Flux TTS) were shown on the pricing page; they are left out, so the prices on this page are list prices.
What to consider when choosing a Deepgram Aura alternative
- Price your monthly characters on each Deepgram Aura tier you would use.
- If you stream text in as a language model produces it, confirm that Deepgram Aura's WebSocket input handles partial sentences the way your code expects.
- Deepgram Aura's per-request limit is not stated on the pages we reviewed, so test your longest text before you commit.
- List the pronunciation problems you have today (names, acronyms, numbers) and test them, since not every service lets you correct them with markup.
Options besides Deepgram Aura
Three other vendors from our dataset, chosen because they share Deepgram Aura's category or pricing model, and ReadAloud, which we make. They are listed alphabetically, not ranked, and each summary comes from that vendor's own pages.
Cartesia
Cartesia is also a text-to-speech API platform, and it lists a price per character as Deepgram Aura does. Its pages mention HIPAA, SOC 2 and GDPR.
ElevenLabs
ElevenLabs is also a text-to-speech API platform, and it lists a price per character as Deepgram Aura does. Its pages mention HIPAA.
ReadAloud
A streaming text-to-speech API: ReadAloud Live at $4 and ReadAloud Studio at $10 per 1M characters. English today.
Typecast API
Typecast API is also a text-to-speech API platform, and it lists a price per character as Deepgram Aura does.
How ReadAloud differs from Deepgram Aura
Deepgram Aura's per-character figures are in the pricing list above, next to ReadAloud Live at $4 and ReadAloud Studio at $10. Live: one American English voice. Studio: several English voices. Requests are limited to 5,000 characters.
Details about Deepgram Aura come from its public pages as they were on October 10, 2026. Deepgram Aura may have changed them since, and a page that was unclear to us may be clear to you. Corrections: support@readaloudai.org. We make ReadAloud, so read this page with that in mind and check anything important with Deepgram Aura directly.
Sources
- Deepgram pricing page (retrieved October 10, 2026)
- Deepgram Flux TTS product page (retrieved October 10, 2026)
- Deepgram: Introducing Aura-2 (retrieved October 10, 2026)
- Deepgram docs: TTS models and languages overview (retrieved October 10, 2026)
- Deepgram docs: TTS media output settings (retrieved October 10, 2026)
- Deepgram docs: API rate limits (retrieved October 10, 2026)
- Deepgram API reference: Single text request (retrieved October 10, 2026)
- Deepgram docs: Migrating from Aura to Flux TTS (retrieved October 10, 2026)
- Deepgram docs: TTS getting started (SDK list) (retrieved October 10, 2026)