Skip to content
Avelto
Pricing

Self-serve pricing. No 6-week onboarding.

Start your team in minutes — not quarters. The cheapest serious option in the AI-roleplay category, with the same multilingual native rubric and audit trail.

Starter

$79per seat / month

For SMB teams up to 50 seats. Card-self-serve, running today.

  • Up to 50 seats
  • 200 conversations / seat / month
  • Default sales + support rubric
  • Multilingual (EN, DE, FR, IT, ES, PL)
  • Manager dashboard + CSV export
  • Email support

Pro

Most popular
$129per seat / month

For scaling orgs (51–300 seats) that need their own rubric and SSO.

  • 51–300 seats
  • 600 conversations / seat / month
  • Custom rubric + brand-voice grounding
  • SSO (SAML / OIDC)
  • Manager override + audit log
  • Per-org prompt versioning
  • Priority support, 24h response

Enterprise

Customcontact us

For 300+ seats with custom DPA, voice mode, and on-call support.

  • Unlimited seats
  • Voice mode (when shipped)
  • Custom DPA and data residency
  • SOC 2 Type II report
  • Calibration loop with our applied AI team
  • Dedicated CSM + on-call
  • Annual or multi-year contracts

Compare plans

Everything in Starter, included in Pro. Everything in Pro, included in Enterprise.

FeatureStarterProEnterprise
Seats includedUp to 5051–300Unlimited
Conversations / seat / month200600Unlimited
Languages (EN, DE, FR, IT, ES, PL)
Default sales + support rubric
Custom rubric
Brand-voice grounding (your playbook)
Per-org prompt versioning + A/B
SSO (SAML / OIDC)
Manager dashboard + CSV
Manager score override
Audit log + 7-yr retention
Anti-cheat signals
Custom DPA
SOC 2 Type II report
Voice mode (post v1.5)
Bring your own LLM keyOptional
SupportEmailPriority 24hDedicated CSM

Frequently asked

  • A conversation is one full roleplay session with the AI persona, from start to either AI-ended, user-ended, or a turn cap. Conversations that end before the second turn are free.

  • Yes. Starter and Pro are month-to-month and cancellable from the workspace billing page. You keep access through the end of the paid period and can export everything via /me/export.

  • No. We never train any model on your conversations or scores. Anthropic and OpenAI are bound by Data Processing Agreements that prohibit training on customer data. Transcripts are encrypted at rest with per-tenant keys.

  • EN, DE, FR, IT, ES, PL ship in v1, with native persona prompts and rubrics for each. UA and RU are on the roadmap.

  • On Pro and Enterprise — yes. Configure your Anthropic or OpenAI key in workspace settings; we route both Persona and Evaluator calls through your key with usage attribution per seat.

Still deciding?

Avelto — Train Sales & Support to Production-Ready