Home / Directory / AI SaaS tooling / Compared / Deepgram vs Cartesia

Deepgram vs Cartesia

Deepgram and Cartesia both sell speech models to teams building voice agents. Deepgram, founded in 2015, started with speech to text and now prices transcription, text to speech, and a full Voice Agent API by usage, with a self-hosted option. Cartesia, founded in 2023, started with fast text to speech and sells monthly credit plans that cover its Sonic voices, Ink transcription, and Line voice agents.

Deepgram vs Cartesia at a Glance, November 2026

Deepgram vs Cartesia at a Glance, November 2026
What buyers compare	Deepgram	Cartesia
Price per month	Pay as you go after a $200 free credit. Growth from $4,000 a year prepaid, up to 20% off. Enterprise: quote	Free: $0. Pro $5, Startup $49, and Scale $299 a month. Enterprise: quote
Free plan or trial	$200 free credit that does not expire, with no card required	Free plan with 20,000 credits and $1 of voice agent use a month
How usage is counted	Per second of audio for transcription, per 1,000 characters for text to speech, and per minute of connection time for voice agents. Nova-3 streaming costs $0.0077 a minute and Aura-2 $0.030 per 1,000 characters	Monthly credits for speech, plus prepaid voice agent dollars on most plans
Integrations and features	Nova-3 and Flux speech to text in more than 60 languages, Aura-2 text to speech, a Voice Agent API, and audio intelligence add-ons	Sonic text to speech, Ink speech to text, voice cloning, and Line for building and running phone and web voice agents
Security and deployment	SOC 2 Type 2, HIPAA BAAs on Enterprise, an EU data endpoint, and self-hosted containers on Enterprise	Scale adds SSO, DPAs, and BAAs. Enterprise adds on-premise or private cloud deployment
Best for	Product teams building production voice agents or transcription that need a specialist speech API	Developers and product teams building real-time voice agents for support, scheduling, sales, or healthcare calls, where the time before the first word of audio decides whether a caller stays on the line

List prices in US dollars per month before tax, as published on each vendor's site on November 6, 2026, billed monthly unless the cell says otherwise. "Quote" means the vendor sets the price through its sales team.

Which one fits

Deepgram is the stronger pick for teams whose voice product starts with transcription at scale and pay-per-second pricing, while Cartesia is the stronger pick for teams whose agents live or die on fast, natural-sounding voices.

Choose Deepgram if

Speech to text is the core of your product, from call transcripts to live captions.
You want to pay per second of audio with no monthly plan.
You need self-hosted speech in your own cloud or data center.

Choose Cartesia if

Your agent's voice quality and speed are what callers notice first.
You want a monthly plan with a set credit budget, from $5 a month.
You want text to speech, transcription, and agent tooling from one smaller vendor.

What buyers say

“The Deepgram worked brilliantly well with transcription across the entire studio, even for un-microphoned guest participants.”

TuringNYC · in a comment on Hacker News · December 18, 2023

“Problem with this is Deepgram's accuracy (but agree their speed/latency is excellent). We used to use them too, but eventually we got so frustrated with poor accuracy we switched to Speechmatics.”

ty00001 · in a comment on Hacker News · December 18, 2023

“Check out Sonic (cartesia.ai). Great quality, very fast - but with a few kinks to work out (going off the rails on long utterances, random sounds, etc).”

nmfisher · in a comment on Hacker News · June 12, 2024

“I'm using a combination of Cartesia and the self-hosted misolabs for TTS. I'm using Deepgram and in some cases Sarvam as well.”

sameersri2004 · describing a voice agent stack in a comment on Hacker News · June 20, 2026

Verdict

Deepgram (8.0 out of 10, Recommend) is the choice for teams that build on transcription at scale, want usage pricing billed by the second, and may need self-hosted speech. Cartesia (7.9 out of 10, Recommend) is the choice for teams building real-time voice agents that care most about fast, natural voices, with monthly plans from $5 and private deployment on Enterprise.

Methodology

We read the Deepgram and Cartesia pricing pages on November 6, 2026 and recorded every published price, limit, and plan feature. A cell reads "Not published" when the vendor does not state that detail on its site.

Scores and best-fit notes come from the AIR reviews of each product. Buyer quotes come from public review sites and vendor-published customer quotes. Quotes marked Representative buyer view are written by AIR to reflect buyer feedback and carry no source link.

How AIR scores vendors.

Sources

Comparison table

Sources by row, numbered as in the list above: Price per month (Sources 3, 4); Free plan or trial (Sources 3, 4); How usage is counted (Sources 3, 4); Integrations and features (Sources 3, 4); Security and deployment (Sources 3, 4); Best for (Sources 1, 2).

Quote sources

“The Deepgram worked brilliantly well with transcription across the…” (TuringNYC · in a comment on Hacker News · December 18, 2023): source
“Problem with this is Deepgram's accuracy (but agree their…” (ty00001 · in a comment on Hacker News · December 18, 2023): source
“Check out Sonic (cartesia.ai). Great quality, very fast -…” (nmfisher · in a comment on Hacker News · June 12, 2024): source
“I'm using a combination of Cartesia and the self-hosted…” (sameersri2004 · describing a voice agent stack in a comment on Hacker News · June 20, 2026): source