74 / 100

Cartesia pricing teardown

https://cartesia.ai/pricing

Cartesia's page is a clean, transparent five-tier ladder with real prices, a usage-based credit allowance, and an unusually good plan-recommendation calculator and FAQ. Weaknesses: no annual billing option, a very thin $5 Pro tier, a dual-allowance value metric (credits plus prepaid agent dollars) that takes real effort to parse, and almost no social proof or trust badges visible near the tiers.

Tier structure

Free

$0/mo

Pro

$5/mo

Startup

$49/mo

Scale

$299/mo

Enterprise

Custom

Value metric

Hybrid: monthly credits plus prepaid agent dollars

How limits scale

Credits / month 20K → 100K → 1.25M → 8M → Custom
Prepaid agent dollars / month $1 → $5 → $49 → $299 → Volume pricing
TTS concurrent requests 2 → 3 → 5 → 15 → Custom
Concurrent voice-agent calls 8 → 12 → 20 → 60 → Custom
Professional voice clones None → None → 2 → 4 → Custom

Escalation logic

Flat monthly subscriptions bundle a credit allowance (TTS/STT) and an equal-dollar prepaid voice-agent balance, with overages on paid plans. Each tier is labeled 'Everything in [previous], plus' and adds licensing, cloning, support, concurrency, and compliance features. Price jumps are steep and uneven: roughly 10x ($5 to $49) and 6x ($49 to $299).

Psychological anchors

  • Free tier as entry anchor — A $0 plan with real allowances (20K credits, $1 of agents) and a 'Start free' CTA lowers the barrier to trying TTS/STT with no trial clock.
  • Highlighted recommended plan — Scale is the only tier with a filled green CTA button. It is the clearest visual emphasis on the page, though there is no 'Most popular' badge and the highlighted tier is the most expensive public one.
  • Enterprise as price anchor — The 'Custom' tier with SSO, DPAs/BAAs, shared Slack, and security questionnaires makes $299 feel moderate and signals enterprise-readiness.
  • Low-cost Pro entry point — $5/mo is a near-impulse purchase that unlocks the commercial-use license and instant cloning, converting free users who need to ship.
  • Equal-dollar prepaid agent credit — Each plan's price returns as agent balance (e.g. $299 plan includes $299 of agent usage), framing the subscription as a spend commitment rather than a cost.
  • Interactive usage calculator — A 'pick your products and volume, we'll recommend a plan' widget reduces plan-selection anxiety and nudges users to the right tier.
  • Strikethrough/promo pricing on agent minutes — The comparison table shows the previous $0.06/min crossed out against the new $0.03/min, adding a value-drop anchor.
  • Charm pricing absent — Prices use $5/$49/$299 with a mix of round and near-charm numbers. $49 is a charm-style price, while $5 and $299 are not consistently styled.

What this page is optimizing for

Primarily self-serve conversion and expansion for developers: public prices, instant 'Start free'/'Select' CTAs, a usage calculator, and overage rates that let usage grow past the plan. Enterprise lead-gen is secondary via the Custom tier, a 'Contact Sales' nav button, and a 'Talk to Cartesia' chat widget.

Red flags

  • No annual billing toggle or discount; every plan is monthly-only, forgoing the typical 17-20% discount and the retention and cash-flow benefits of annual commitments.
  • Value metric is dual (credits for TTS/STT plus prepaid dollars for agents), and users must do math (1 credit per character, 3 credits per second, ~$0.03/min) to understand what a plan really buys.
  • Price steps are large and uneven (about 10x from Pro to Startup, 6x from Startup to Scale), so mid-sized users face an abrupt jump or must rely on overages.
  • Five tiers is at the upper edge of what's easy to compare, and the two lowest tiers ($0 and $5) differ mainly by license and cloning, not by a clear user type.
  • Key limits (concurrency, number of phone numbers, voice-clone slots) sit in the detailed comparison table and FAQ rather than on the tier cards, so tier differences on the cards are thin.
  • The screenshot shows no 'Most popular' badge or other label explaining why Scale is highlighted; there are also no customer logos, testimonials, or review scores near the pricing cards.
  • Overages are off by default on paid plans and unavailable on Free, so an unexpected cutoff of calls or generation is possible, and Pro/Startup/Scale overage rates differ ($65/$45/$38 per 1M credits).

Best-practices scorecard

What works · 3

  • Visible prices — Four of five tiers show exact monthly prices, and overage rates and per-minute costs are published in the FAQ.
  • FAQ or objection handling — An extensive FAQ covers rollover, upgrades, downgrades, overages, credit math, concurrency, and security, plus a usage calculator for plan sizing.
  • Clear CTAs per tier — Each tier has a specific action CTA ('Start free', 'Select Pro', 'Contact us'), and a persistent 'Talk to Cartesia' widget supports questions.

Half measures · 4

  • Clear recommended tier — Scale has the only filled CTA, but there is no badge or label explaining the recommendation, and the highlighted plan is the highest public price.
  • Value metric matches usage — Credits and per-minute agent pricing track consumption well, but the two-currency model (credits plus prepaid dollars) adds cognitive load.
  • Tier differences are scannable — 'Everything in X, plus' lists are tidy, but the cards show only two bullets of new features and the real differences live in a separate detailed comparison table.
  • Trust signals present — SOC 2 Type II, DPAs/BAAs, and SSO are mentioned in the Enterprise tier and FAQ, but there are no customer logos or testimonials near the pricing cards.

What's missing · 1

  • Annual discount offered — No annual billing option or discount is shown anywhere; all plans are monthly.