Annual subscriptions are billed once per year. Under the current Early Access billing implementation, the usage allowances shown below apply to the full annual billing period rather than resetting monthly.
These rates apply to the Basic, Pro, and Business plans. Enterprise capacity and overage rates are negotiated.
| AI mode | Call | Voice minute | SMS initiation | Email initiation |
|---|---|---|---|---|
| Platform managed | $0/call | $0.35/min | $0.025/SMS initiation | $0.005/email initiation |
| Bring your own LLM | $0/call | $0.12/min | $0.015/SMS initiation | $0.003/email initiation |
Voice calls, two-way SMS, and inbound or outbound email. Each workflow uses one channel and one destination; the same agent configuration can be reused in separate channel-specific workflows.
Configure an agent's role, objectives, tone, language hints, opening and closing scripts, topic rules, guardrail settings, and conversation limits. Knowledge-base grounding is available on Pro and higher plans.
Platform-managed AI currently uses OpenAI. For text workflows, you can bring keys for OpenAI, Azure OpenAI, Google Gemini, Anthropic, or a custom OpenAI-compatible endpoint. Realtime voice with your own key supports OpenAI and Azure OpenAI. AWS Bedrock is coming soon.
Packaged CRM and HR connectors are not available today. Contact us about the systems and data flows your team needs so we can assess an integration path.
Configurable rules can evaluate text and transcripts for sensitive values, blocked topics, toxicity, opt-out language, and conversation limits. Events are recorded and can flag or halt a conversation. Realtime voice audio reaches the selected provider before transcript-level controls can run.
After a conversation completes, asynchronous analysis can produce summaries, sentiment, emotion, intent, frustration, engagement, language-quality, topic, outcome, agent-quality, and guardrail signals. Signals can degrade or abstain when the evidence is insufficient.
User access uses RS256-signed tokens, refresh-token revocation, ownership and user/admin/system role checks, and route-level rate limits. Internal service calls use service credentials, while Twilio, SendGrid, and Stripe callbacks are verified at their respective boundaries.
Each tier has a base subscription and included allowances for calls, voice minutes, and outbound SMS and email initiations. Usage above those allowances is metered at the listed rate. Monthly subscriptions use a monthly usage period. Under the current Early Access billing implementation, annual-plan allowances apply once across the annual billing period rather than resetting monthly. Platform-managed and bring-your-own-LLM modes have different base prices.
Each initiated call counts toward the call allowance; each completed call’s duration is rounded up to the next whole Voice Minute. A successful initial outbound SMS or email workflow send counts once toward its respective allowance. Later conversation replies do not currently add SMS or email usage units. Usage events are aggregated for capacity reporting and billing-cycle close.
Yes. Eligible account holders can open Stripe's Billing Portal to manage a subscription. Effective dates and available changes depend on the active subscription and Stripe configuration.
The platform stores conversation text, transcripts, metadata, guardrail events, analysis results, and usage references needed to operate the service. Access controls apply at user and service boundaries; sensitive text values are redacted before supported persistence and egress paths.
Chatter Genius is in early access. Timing depends on workspace approval, sender provisioning, model-provider setup, content review, and workflow testing. Contact us to plan a realistic launch path.
