Plans and how credits work

Wassla offers three paid plans — Starter at $49, Growth at $199, and Scale at $399 per month — and meters AI work in credits, where one credit equals one US cent (USD 0.01) of spend value. Every workspace also runs in one of three billing modes — infra_only, demo, or live — which controls whether AI replies are actually paid for. This page explains what each plan includes, what consumes credits, and how the billing modes interact with your monthly allowance.

The three plans at a glance

Wassla pricing is usage-based, not seat-based. You can invite the whole team on any tier — what you pay for is the channels you connect, the volume of conversations the AI handles, and the monthly credit allowance bundled in.

Starter — $49 per month

Starter is the entry tier for teams that want AI on text channels only. It includes the web widget and Twilio SMS, supports up to 2 AI agents, 500 conversations per month, and 200 knowledge entries. The plan ships with 5,000 credits included each month and removes the "powered by Wassla" badge from your widget. Voice is not available on Starter.

Growth — $199 per month

Growth is the most common tier for production deployments. It adds WhatsApp Business, Instagram DMs, Facebook Page Messenger, WhatsApp Calling, and voice agents on the web widget. You get up to 5 AI agents, 2,000 conversations per month, 500 knowledge entries, evals, analytics export, and a custom voice picker. Growth includes 25,000 credits per month and is the first tier where you can run a full omnichannel AI customer support operation.

Scale — $399 per month

Scale is the production tier for high-volume teams and regulated industries. It includes everything in Growth plus Twilio Voice (inbound and outbound phone trunks), SSO, audit-log export, a contractual SLA, and a dedicated success manager. The plan supports up to 25 AI agents, 10,000 conversations per month, and 2,000 knowledge entries. It ships with 60,000 credits per month. Annual billing on Growth and Scale gets two months free.

For larger deployments — custom voice training, isolated regions, white-label, or enterprise procurement — contact sales at [email protected] for a custom Scale-tier contract.

What one credit is worth

One Wassla credit equals one US cent (USD 0.01) of AI spend value. The plan grant is real money you can use: 25,000 credits on Growth equals $250 of bundled AI work each month, on top of the $199 platform fee. The credit accounting layer prices each unit of work in credits per unit, then debits your workspace balance in real time as the AI runs.

This pricing model exists for two reasons. First, it lets you reason about AI cost the same way you reason about Twilio segments or WhatsApp template messages — as a discrete unit you can budget against. Second, it gives Wassla a single internal currency that works across LLM providers, speech-to-text vendors, embeddings, and voice minutes without surfacing provider math to you.

What activities consume credits

Every AI-driven action consumes credits. Human actions in the Inbox — your teammates taking over a conversation, sending a manual reply, tagging a customer — are always free.

The three metered activities are:

  • AI text replies — 2 credits per reply. This covers the LLM call that drafts and sends the response on WhatsApp, Instagram, Facebook, SMS, or the web widget.
  • Voice minutes — 20 credits per minute of active voice conversation. This bundles the speech-to-text turn, the LLM reasoning, and the text-to-speech synthesis for both web-widget voice and WhatsApp Calling.
  • Outbound messages — 2 credits per outbound proactive message (campaigns, broadcasts, follow-ups initiated by an AI agent).

Knowledge-base searches, sentiment analysis, embeddings generation, and tool calls are billed through the same metric system but are typically rolled into the cost of the reply they support. See How Wassla credits work for the live per-transaction ledger.

The three billing modes

Wassla has three billing modes that control whether AI spend is actually permitted on your workspace. Every workspace starts in infra_only and you choose when to flip to demo or live.

infra_only (the default)

In infra_only mode, AI replies are turned off platform-wide for your workspace. You can still connect every channel, train AI agents, upload knowledge, invite teammates, and configure routing — but no LLM, embedding, or voice call will be made. Inbound messages land in the Inbox for human handling. This mode is the safe default for setup and integration work, and it costs nothing in AI spend. A pre-live banner appears across the workspace while you are in this mode.

demo

In demo mode, AI replies are turned on but capped by a monthly demo allowance. Once the allowance is exhausted, AI replies pause until the next month or until you upgrade to live. Demo mode is the right choice for evaluation, internal pilots, and Wassla-led demos — it lets your agents respond end-to-end without committing to a full plan. Demo spend is tracked against the same credit ledger as live spend, and a TOCTOU-safe reservation system prevents concurrent requests from overshooting the monthly cap.

live

In live mode, AI replies are fully enabled and metered against your plan's monthly credit grant. When the grant is exhausted, you either top up, upgrade, or AI replies pause and new conversations queue for human handling. Live is the production mode — switch to it the day you go live with real customers.

The billing-mode resolver runs server-side on every AI call and fails closed: if it cannot resolve the mode it defaults to infra_only, so a misconfigured workspace can never accidentally spend. Tenant-level overrides beat the platform default, and an Owner can flip modes from the Billing page (with dual-control approval on platform-wide changes).

How plan grants, credits, and billing mode work together

The three layers compose like this. Your plan decides which channels and features are unlocked and how many credits you get each month. Your credit balance is what you spend as AI actions run. Your billing mode decides whether spend is actually permitted right now. If billing mode is infra_only, no credits are spent regardless of balance. If billing mode is demo or live and the balance reaches zero, AI replies pause until you top up or the next monthly grant lands.

You can monitor all three from a single screen: open Billing in the workspace sidebar. The dashboard shows your active plan, your current credit balance, this month's spend by metric, and the live billing mode for your workspace.