Cozzo
Find yours

Cartesia Pricing Explained (2026): Credits, Sonic TTS Minutes and Voice Agent Costs

Cartesia has two kinds of pricing. Its Sonic text-to-speech and Ink speech-to-text models are paid for with monthly credits that come with a plan: Free (20K credits), Pro at $5 a month (100K), Startup at $49 (1.25M) and Scale at $299 (8M), with Enterprise custom. Its hosted Managed Agents are billed separately in dollars, at $0.06 a minute, and every plan's fee also comes back as prepaid agent dollars. Every figure below is from cartesia.ai/pricing (plans, comparison table and FAQ) and Cartesia's pricing docs, read October 22, 2026.

Cartesia prices speech in credits, and one credit is one character. Think of a minute of Sonic speech as about 750 credits, and every plan becomes a simple number of minutes.

Cartesia sells monthly credit plans from free to $299, and a minute of Sonic speech takes about 750 to 800 credits. Here's what each plan really buys in minutes of speech and hours of Ink transcription, the overage rates, the $0.06/min Managed Agents price, and where Startup or Scale starts paying off.

What a Cartesia voice agent minute costs

  • Managed Agents call time

    $0.06/min

    The same on Free, Pro, Startup and Scale, billed in dollars, not credits. Each plan includes prepaid agent dollars equal to its fee ($1 a month on Free), about 17 minutes on Free, 1 hour 23 minutes on Pro, 13 hours 37 minutes on Startup and 83 hours on Scale. Past that you keep paying $0.06 a minute.

  • Phone line

    $0–$0.014/min

    Cartesia adds $0.014 a minute when the call uses a Cartesia-provided phone number (1 number on Free, 3 on Pro, 5 on Startup, 10 on Scale).

  • LLM

    Per token

    Not included in the per-minute rate. Token prices depend on the model your agent uses, and Cartesia's API lists them per model. Its pricing page says LLM usage during calls is free for a limited time on agents created in Cartesia's UI.

  • Total per minute

    about 6.0¢ to 7.4¢

    A Managed Agents call minute before LLM tokens. If you use Sonic or Ink inside your own voice stack instead, you pay in credits from your plan, as in the table below.

What a month of Cartesia text-to-speech costs

  • 500 min of speech a month

    Pro ($5)
    $22.88
    Startup ($49)
    $49.00
    Scale ($299)
    $299.00

  • 1,000 min of speech a month

    Pro ($5)
    $47.25
    Startup ($49)
    $49.00
    Scale ($299)
    $299.00

  • 2,500 min of speech a month

    Pro ($5)
    $120.38
    Startup ($49)
    $77.13
    Scale ($299)
    $299.00

  • 5,000 min of speech a month

    Pro ($5)
    $242.25
    Startup ($49)
    $161.50
    Scale ($299)
    $299.00

  • 10,000 min of speech a month

    Pro ($5)
    $486.00
    Startup ($49)
    $330.25
    Scale ($299)
    $299.00

  • 25,000 min of speech a month

    Pro ($5)
    $1,217.25
    Startup ($49)
    $836.50
    Scale ($299)
    $707.50

  • 50,000 min of speech a month

    Pro ($5)
    $2,436.00
    Startup ($49)
    $1,680.25
    Scale ($299)
    $1,420.00

Minutes of generated Sonic speech, not call minutes. Each plan's fee plus overage for anything past its credits, at 750 credits a minute (the figure Cartesia's own plan table uses; its FAQ says 750 to 800). Included speech is about 133 minutes on Pro, 1,667 on Startup and 10,667 on Scale; overage, which you have to switch on, is $65 per 1M credits on Pro, $45 on Startup and $38 on Scale (about 4.9¢, 3.4¢ and 2.9¢ a minute of speech). Rolled-over credits and Ink speech-to-text usage aren't counted. Want your own volume and carrier? Try the cost calculator.

Packages and add-ons

  • Free

    $0

    20K credits a month (about 27 minutes of Sonic speech or 1 hour 51 minutes of Ink-2 transcription) and $1 of prepaid agent dollars. 2 concurrent text-to-speech requests, 8 speech-to-text and 8 agent calls, and 1 Cartesia phone number. Cartesia lists the commercial use license from Pro up.

  • Pro

    $5/mo

    100K credits (about 133 minutes of speech or 9 hours 16 minutes of Ink-2) and $5 of agent dollars. Adds the commercial use license and instant voice cloning. 3 concurrent text-to-speech requests, 12 speech-to-text and 12 agent calls, 3 phone numbers. Overage $65 per 1M credits.

  • Startup

    $49/mo

    1.25M credits (about 1,667 minutes of speech or 115 hours 44 minutes of Ink-2) and $49 of agent dollars. Adds professional voice cloning (2 slots) and organizations. 5 concurrent text-to-speech requests, 20 speech-to-text and 20 agent calls, 5 phone numbers. Overage $45 per 1M credits.

  • Scale

    $299/mo

    8M credits (about 10,667 minutes of speech or 740 hours 44 minutes of Ink-2) and $299 of agent dollars. Adds priority support, 4 professional voice clone slots and higher concurrency: 15 text-to-speech requests, 60 speech-to-text and 60 agent calls, 10 phone numbers. Overage $38 per 1M credits.

  • Enterprise

    Custom

    Custom credits, agent usage, volume pricing and concurrency limits, plus DPAs and BAAs for compliance, SSO, a shared Slack channel and security questionnaires. Cartesia says on-premise or virtual private cloud deployment is available through sales.

  • Ink speech-to-text

    1–3 credits/sec

    Ink-2 is 3 credits per second of audio (10,800 an hour) on the streaming and turn-detection endpoints; batch Ink-2 isn't available yet. Ink-Whisper is 1 credit per second streaming and 1 credit per 2 seconds in batch. Silence is billed too. On Scale's included credits Ink-2 works out to about $0.40 an hour (Cartesia's FAQ rounds it to $0.39).

  • Extras billed in credits

    Per request

    Infill (audio that bridges two clips) is 300 credits plus about 1 credit per character. Localizing a voice is 225 credits per added accent. Each break tag counts as 1 credit. Training a professional voice clone is free and its speech costs the normal 1 credit per character. Agent evaluations are free for a limited time.

  • Rollover and billing rules

    Up to 2x plan

    Unused credits roll over up to twice your monthly allotment, and you keep paid credits when you change plans. Only successful requests use credits. Overages are a toggle: with it off, requests fail once credits run out. Plans renew monthly from the purchase date, and a downgrade takes effect at the end of the period.

How it compares on published pricing

  • Cartesia logoCartesia
    Starting price
    Free; paid from $5/mo
    Best for
    Text-to-speech from a vendor focused on low latency (its own figures)
  • ElevenLabs Agents logoElevenLabs Agents
    Starting price
    Agents: $0.08 per minute
    Best for
    Builders who want ElevenLabs' voices and speech models in one agent pipeline
  • Rime logoRime
    Starting price
    From $0.03/min TTS (Mist v3); STT free until Nov 19, then $0.002/min
    Best for
    Low-latency TTS for phone agents, now with speech-to-text from the same vendor
  • Deepgram logoDeepgram
    Starting price
    Nova-3 streaming $0.0048/min ($0.29/hr); $200 free credit
    Best for
    Streaming speech-to-text for voice agents
  • Speechmatics logoSpeechmatics
    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

Pricing explained, line by line:ElevenLabs API pricing →Deepgram pricing →

TTS models compared

See all voice models · Latency and prices as each vendor publishes them.

What to check before you choose

  • Convert credits to minutes before you pick a plan. A minute of Sonic speech takes about 750 to 800 credits, so Pro ($5) buys about 133 minutes, roughly 3.8¢ a minute, Startup ($49) about 1,667 minutes (2.9¢) and Scale ($299) about 10,667 minutes (2.8¢).
  • Know where the plans cross. On Cartesia's overage rates, 1,000 minutes of speech a month costs about $47 on Pro against $49 on Startup; past about 1,040 minutes Startup is cheaper, and from about 9,100 minutes Scale beats Startup.
  • Concurrency may choose the plan for you. Text-to-speech allows 2, 3, 5 and 15 simultaneous requests on Free, Pro, Startup and Scale. If more than 5 callers could hear your agent speak at the same moment, you need Scale or Enterprise whatever your volume. Speech-to-text and agents have their own, larger pools (8 to 60).
  • The free plan is for testing. Cartesia lists the commercial use license from Pro ($5 a month), along with instant voice cloning, so plan on at least Pro for anything customers hear.
  • Compare per 1,000 characters. Cartesia is about $0.050 on Pro, $0.039 on Startup and $0.037 on Scale (overage $0.065, $0.045 and $0.038). On their own pricing pages, ElevenLabs v4 Turbo and Flash are $0.04, Deepgram Aura-2 is $0.030 ($0.027 on Growth), Rime Mist v3 is $0.03 and Speechmatics is $0.011 with the first 1M characters free.
  • Compare the agent minute fairly. Cartesia Managed Agents is $0.06 a minute plus $0.014 on a Cartesia number, with the LLM extra (free for now on UI-built agents). ElevenLabs Agents is $0.08 a minute plus the LLM, and AssemblyAI's Voice Agent API is $0.075 a minute with the LLM included but the phone carrier extra.
  • Silence counts on Ink. Speech-to-text is billed on audio duration, silence included, so close the stream when the caller hangs up. Failed requests don't use credits.

Questions people ask

How much does Cartesia cost?

Cartesia has a free plan with 20K credits a month and paid plans at $5 (Pro, 100K credits), $49 (Startup, 1.25M) and $299 a month (Scale, 8M), plus custom Enterprise. Sonic text-to-speech uses about 1 credit per character, roughly 750 to 800 credits per minute of speech, so Pro covers about 133 minutes. Its hosted Managed Agents cost $0.06 a minute, plus $0.014 a minute on a Cartesia phone number, and each plan's fee also comes back as prepaid agent dollars.

How many Cartesia credits does a minute of speech take?

About 750 to 800. Sonic text-to-speech costs about 1 credit per character, and Cartesia's FAQ says a minute of generated audio needs 750 to 800 credits. Its plan table works out to about 27 minutes on Free, 133 on Pro, 1,667 on Startup and 10,667 on Scale.

What does Cartesia speech-to-text cost?

Ink-2 uses 3 credits per second of audio (10,800 an hour) on the streaming and turn-detection endpoints, and Ink-Whisper uses 1 credit per second streaming or 1 credit per 2 seconds in batch. Silence is billed. Cartesia's FAQ puts an hour of Ink-2 at about $0.39 on the Scale plan.

What happens if I run out of Cartesia credits?

On a paid plan you can turn on overages and keep going at $65 per 1M credits on Pro, $45 on Startup or $38 on Scale, billed on top of your subscription. With overages off, requests fail until your plan renews or you upgrade.

Do unused Cartesia credits roll over?

Yes, up to twice your monthly allotment. You keep credits you've paid for when you upgrade or downgrade; the rollover cap resets to 2x the new plan's monthly amount.

How much do Cartesia voice agents cost?

Managed Agents are $0.06 a minute on every plan, billed in dollars rather than credits, plus $0.014 a minute when the call uses a Cartesia-provided number. LLM tokens are extra and depend on the model, though Cartesia says LLM usage is free for a limited time on agents built in its UI. Each plan includes prepaid agent dollars equal to its fee, about 83 hours of calls on Scale.

Does Cartesia sign a BAA?

Cartesia lists DPAs and BAAs for compliance on its Enterprise plan, which is custom priced, and says it is SOC 2 Type II certified. Self-serve plans don't list a BAA.

Ready to look closer? See Cartesia’s full listing and price history → Or weigh the options in Cartesia Alternatives for Voice Agent TTS Compared.

Not sure which fits? Answer a few quick questions in the finder →

ElevenLabs Agents vs Vapi →Deepgram vs AssemblyAI →Cartesia vs ElevenLabs →Deepgram vs Speechmatics →Rime vs Cartesia →Deepgram vs ElevenLabs →Retell AI vs ElevenLabs Agents →

Prices from vendors' official pricing pages, checked October 22, 2026. Not sponsored; order isn't a ranking.