# Pricing — HAPP AI

> The plan buys the concurrent lines and the support; everything the assistant does is on every plan. Work is billed at your plan’s rate: voice from $0.022 a minute of conversation, chat from $0.006 a message, analysis from $0.004 a message of the conversation, transcription from $0.003 a minute. AI models are separate and go at the provider’s price plus 20% for taxes and currency movement — we do not resell them. Both halves are on this page, in full.

- **Page:** https://happ.tools/pricing
- **Billing:** A monthly platform subscription, plus the minutes voice agents spend on the line at the model’s own rate
- **Trial:** Free trial on every plan · Nothing charged for 7 days · Cancel any time
- **Action:** [Open HAPP Dashboard](https://my.happ.tools/)

## Plans

*Pricing*

Every feature on every plan — what the plan buys is concurrent lines and how fast we answer you. Work is billed at your plan’s rate: voice from $0.022 a minute of conversation, chat from $0.006 a message, analysis from $0.004 a message. AI models are separate, at the provider’s price — we do not resell them.

| Plan | Price | For | Includes |
| --- | --- | --- | --- |
| Start | $49/month | Small teams | 3 concurrent calls; All channels included; Email support, 24 hours |
| Business | $199/month | Best for business | 7 concurrent calls; SSO · audit log; Telegram manager, 8 hours |
| Scale | $499/month | High volume | 15 concurrent calls; 99.9% SLA · full history; Named manager, 1 hour |
| Enterprise | Custom | Built for large organizations | Negotiated per-minute rate; Dedicated server · SLA; Custom integrations · on-premise |

**Enterprise** — Custom pricing, a dedicated server, SLAs, AI tuned to your processes, on-premise.

**Pay upfront — get a discount** — Choose your billing period at checkout — the discount applies automatically. −8% 3 months · −12% 6 months · −22% 12 months · +bonus on top-up.

## What you get, line by line

*Every feature, every plan*

The same list the plan cards summarise, in full. Everything an agent does is on every plan — what moves with the plan is how many people, channels and lines it runs for, and how deep the analytics and the administration go.

### Team & channels

| Feature | Start | Business | Scale | Enterprise |
| --- | --- | --- | --- | --- |
| Manager seats — *Not a lever any more — a plan is priced on the work the assistant does, not on how many people watch it.* | Unlimited | Unlimited | Unlimited | Unlimited |
| Messenger connections — *Telegram, Instagram, WhatsApp, Viber, Facebook and the rest.* | Unlimited | Unlimited | Unlimited | Unlimited |
| Telephony lines | Unlimited | Unlimited | Unlimited | Unlimited |
| Analytics monitors — *A saved view of one team, funnel or channel.* | Unlimited | Unlimited | Unlimited | Unlimited |
| Concurrent calls — *What the plan actually buys, together with support: how many conversations run at the same moment.* | 3 | 7 | 15 | Negotiated |
| Every channel in one inbox | Yes | Yes | Yes | Yes |
| iOS mobile app | Yes | Yes | Yes | Yes |

### Service fee — our part, next to the model

| Feature | Start | Business | Scale | Enterprise |
| --- | --- | --- | --- | --- |
| Voice, per minute of conversation — *What we charge for running the call. The model is billed separately at the provider’s own price, plus 20% for taxes and currency movement; connect your own key and you skip the 20%.* | $0.045 | $0.032 | $0.022 | Negotiated |
| Message, per message — *Every message the assistant answers, in any messenger.* | $0.012 | $0.009 | $0.006 | Negotiated |
| Analysis, per message — *Per message of the conversation the analysis reads — a chat message or a reply in a call transcript — so a long call costs more than a short chat. The model is charged on top by its tokens: about $0.0004 per dialog on our own analyst, from $0.001 on an external one.* | $0.006 | $0.005 | $0.004 | Negotiated |
| Transcription, per minute of recording — *For recordings that did not go through our voice agent. It runs on our own servers, so there is no model line at all — and $0.005 a minute undercuts AssemblyAI, Whisper and Deepgram.* | $0.005 | $0.004 | $0.003 | Negotiated |
| AI models — *Pick any model in the catalogue. You pay the provider price plus 20% for taxes and currency movement — we take no cut on the model itself. Bring your own key and the 20% goes away.* | Provider price | Provider price | Provider price | Provider price |

### The AI agent

| Feature | Start | Business | Scale | Enterprise |
| --- | --- | --- | --- | --- |
| Voice agent on your phone lines | Yes | Yes | Yes | Yes |
| Chat agent on every connected messenger | Yes | Yes | Yes | Yes |
| Any language, switched mid-conversation | Yes | Yes | Yes | Yes |
| Answers only from your knowledge base — *When something is not in it, the agent says so instead of inventing an answer.* | Yes | Yes | Yes | Yes |
| Acts in your CRM — orders, bookings, deal updates | Yes | Yes | Yes | Yes |
| Hands over to a manager with the whole thread | Yes | Yes | Yes | Yes |
| Custom voice, tone and escalation rules | Yes | Yes | Yes | Yes |
| Any model on the catalogue — *The model is a setting on the agent, not a plan.* | Yes | Yes | Yes | Yes |

### Conversation analytics

| Feature | Start | Business | Scale | Enterprise |
| --- | --- | --- | --- | --- |
| Transcription of every call and chat | Yes | Yes | Yes | Yes |
| Sentiment analysis | Yes | Yes | Yes | Yes |
| Manager scoring | Yes | Yes | Yes | Yes |
| Scoring templates & key moments | Yes | Yes | Yes | Yes |
| Weekly email reports | Yes | Yes | Yes | Yes |
| Google Sheets export | Yes | Yes | Yes | Yes |
| Daily reports | Yes | Yes | Yes | Yes |
| Advanced analytics & manager comparison | Yes | Yes | Yes | Yes |
| Conversation history | 3 months | 24 months | Unlimited | Unlimited |

### Integrations & API

| Feature | Start | Business | Scale | Enterprise |
| --- | --- | --- | --- | --- |
| Messengers, telephony and CRM out of the box | Yes | Yes | Yes | Yes |
| Call & chat webhooks | Yes | Yes | Yes | Yes |
| MCP server — *Point Claude, Cursor or your own agent at your HAPP workspace.* | Yes | Yes | Yes | Yes |
| REST API access | — | — | Yes | Yes |
| Integrations built for you | — | — | — | Yes |

### Security & administration

| Feature | Start | Business | Scale | Enterprise |
| --- | --- | --- | --- | --- |
| GDPR, SOC 2 Type 2 and ISO 27001 | Yes | Yes | Yes | Yes |
| 2FA with email codes | Yes | Yes | Yes | Yes |
| Advanced permissions | Yes | Yes | Yes | Yes |
| Custom user roles & flexible groups | Yes | Yes | Yes | Yes |
| SAML single sign-on | — | — | Yes | Yes |
| Audit logging | — | — | Yes | Yes |
| User provisioning | — | — | Yes | Yes |
| Regional deploys | — | — | — | Yes |
| Dedicated server or on-premise | — | — | — | Yes |

### Onboarding & support

| Feature | Start | Business | Scale | Enterprise |
| --- | --- | --- | --- | --- |
| Free 7-day trial | Yes | Yes | Yes | Yes |
| Live in a day, with docs and a knowledge base | Yes | Yes | Yes | Yes |
| First response time | Email | 4 hours | 1 hour | Negotiated |
| Uptime SLA | — | 99.5% | 99.9% | Negotiated |
| Named success manager | — | — | Yes | Yes |

Limits are per workspace. Need one number higher than the plan you are on? Ask — we raise a single limit rather than sell you the next plan.

## AI models and what a minute of talking costs

*Pay per minute*

On top of the subscription and the service fee you pay the model itself, at the provider’s price plus 20% for taxes and currency movement. We take no cut on the model: bring your own provider key and the 20% goes away. Move an agent to a cheaper model at any point — it is a setting, not a plan.

| Model | Provider | Modalities | Per minute |
| --- | --- | --- | --- |
| GPT Realtime 2 | OpenAI | audio + text | $0.1360/min |
| GPT Realtime 1.5 | OpenAI | audio + text | $0.1360/min |
| Gemini 3.1 Flash Live (preview) | Google Gemini | audio + text | $0.0667/min |
| Gemini 2.5 Flash Native Audio (latest) | Google Gemini | audio + text | $0.0667/min |
| Grok Voice (latest) | xAI | audio + text | $0.0800/min |
| Grok Voice Think Fast 1.0 | xAI | audio + text | $0.0500/min |
| gpt-3.5-turbo | ElevenLabs | text | $0.0733/min |
| gpt-4-turbo | ElevenLabs | text | $0.2509/min |
| gpt-4o | ElevenLabs | text | $0.1115/min |
| gpt-4o-mini | ElevenLabs | text | $0.0668/min |
| gpt-4.1 | ElevenLabs | text | $0.1020/min |
| gpt-4.1-mini | ElevenLabs | text | $0.0716/min |
| gpt-4.1-nano | ElevenLabs | text | $0.0659/min |
| gpt-5 | ElevenLabs | text | $0.0892/min |
| gpt-5-mini | ElevenLabs | text | $0.0690/min |
| gpt-5-nano | ElevenLabs | text | $0.0650/min |
| gpt-5.1 | ElevenLabs | text | $0.0892/min |
| gpt-5.2 | ElevenLabs | text | $0.0993/min |
| gpt-5.4 | ElevenLabs | text | $0.1179/min |
| gpt-5.4-mini | ElevenLabs | text | $0.0802/min |
| gpt-5.4-nano | ElevenLabs | text | $0.0683/min |
| gpt-5.5 | ElevenLabs | text | $0.1718/min |
| gpt-5.6-sol | ElevenLabs | text | $0.1718/min |
| gpt-5.6-terra | ElevenLabs | text | $0.1071/min |
| gpt-5.6-luna | ElevenLabs | text | $0.0683/min |
| gemini-2.0-flash | ElevenLabs | text | $0.0659/min |
| gemini-2.0-flash-lite | ElevenLabs | text | $0.0654/min |
| gemini-2.5-flash | ElevenLabs | text | $0.0668/min |
| gemini-2.5-flash-lite | ElevenLabs | text | $0.0659/min |
| gemini-3.1-flash-lite | ElevenLabs | text | $0.0689/min |
| gemini-3.5-flash | ElevenLabs | text | $0.0934/min |
| gemini-3.5-flash-lite | ElevenLabs | text | $0.0701/min |
| claude-3-haiku | ElevenLabs | text | $0.0688/min |
| claude-3-5-sonnet | ElevenLabs | text | $0.1219/min |
| claude-3-7-sonnet | ElevenLabs | text | $0.1219/min |
| claude-haiku-4-5 | ElevenLabs | text | $0.0833/min |
| claude-sonnet-4 | ElevenLabs | text | $0.1219/min |
| claude-sonnet-4-5 | ElevenLabs | text | $0.1219/min |
| claude-sonnet-4-6 | ElevenLabs | text | $0.1219/min |
| claude-sonnet-5 | ElevenLabs | text | $0.1064/min |
| claude-opus-4-7 | ElevenLabs | text | $0.1605/min |
| claude-opus-4-8 | ElevenLabs | text | $0.1701/min |
| qwen3-4b | ElevenLabs | text | $0.0698/min |
| qwen3-30b-a3b | ElevenLabs | text | $0.0698/min |
| qwen35-397b-a17b | ElevenLabs | text | $0.0797/min |
| qwen36-35b-a3b | ElevenLabs | text | $0.0678/min |
| glm-45-air-fp8 | ElevenLabs | text | $0.0836/min |
| watt-tool-8b | ElevenLabs | text | $0.0669/min |
| watt-tool-70b | ElevenLabs | text | $0.0669/min |
| GPT-4o (OpenAI) | Deepgram | text | $0.0750/min |
| GPT-4o mini (OpenAI) | Deepgram | text | $0.0750/min |
| GPT-4.1 (OpenAI) | Deepgram | text | $0.0750/min |
| GPT-4.1 mini (OpenAI) | Deepgram | text | $0.0750/min |
| GPT-5 (OpenAI) | Deepgram | text | $0.0750/min |
| GPT-5.4 (OpenAI) | Deepgram | text | $0.0750/min |
| GPT-5.4 mini (OpenAI) | Deepgram | text | $0.0750/min |
| GPT-5.5 (OpenAI) | Deepgram | text | $0.0750/min |
| Claude 3.5 Haiku (Anthropic) | Deepgram | text | $0.0750/min |
| Claude Haiku 4.5 (Anthropic) | Deepgram | text | $0.0750/min |
| Claude Sonnet 4.5 (Anthropic) | Deepgram | text | $0.0750/min |
| Claude Sonnet 4.6 (Anthropic) | Deepgram | text | $0.0750/min |
| Gemini 2.0 Flash Lite (Google) | Deepgram | text | $0.0750/min |
| Gemini 2.5 Flash (Google) | Deepgram | text | $0.0750/min |
| Gemini 3 Pro (preview) (Google) | Deepgram | text | $0.0750/min |
| Gemini 3.1 Flash Lite (Google) | Deepgram | text | $0.0750/min |
| Gemini 3.5 Flash (Google) | Deepgram | text | $0.0750/min |
| Nemotron 3 Nano (NVIDIA) | Deepgram | text | $0.0750/min |

## Frequently asked questions

*FAQ*

What the subscription covers, how minutes are counted, and what happens at a limit.

### What does the monthly price cover?

Access to the platform and everything in it — unlimited seats, channels, telephony lines and agents, the inbox and the conversation analytics. It does not cover the work the assistant does: minutes, messages, analyses and transcription are billed per unit at your plan’s rate, and model tokens are passed through at the provider’s own price. The model line carries 20% for taxes and currency movement when it is charged; connect your own key and you skip it.

### How are AI minutes counted?

Per minute the agent is on the line, at your plan’s rate — $0.045 on Start, $0.032 on Business, $0.022 on Scale. Messages the assistant answers and messages the analysis reads are counted the same way, per message. The model’s own tokens are billed separately at the provider’s own price, so a cheaper model really is cheaper for you. The model line carries 20% for taxes and currency movement when it is charged; connect your own key and you skip it. Idle time is never charged. Both come out of a balance you top up, so a busy month costs more and a quiet one costs less.

### Can I change the model later?

Yes, and it takes one setting. The model belongs to the agent, not to the plan — switch it and the new rate applies from the next call, on any plan.

### Can I change plan later?

Move up whenever you need the headroom; the difference is prorated against what you have already paid. Moving down takes effect at the start of the next billing period.

### What happens when I hit a limit?

Nothing stops answering. You are told which limit you reached and can add what you need — a seat, a line, a channel — or move up a plan. Customers do not see a queue because of your billing.

### Is there a free trial?

Yes, on every plan. Full feature access for 7 days — you connect a channel and watch the agent answer before anything is charged.

### Do you invoice companies?

Yes, with the closing documents your accountant expects. Paying for 3, 6 or 12 months up front takes 8, 12 or 22 percent off, and the discount applies automatically at checkout.

### When is Enterprise the right answer?

When the limits are not the problem — a dedicated server, an SLA, on-premise, regional deploys, integrations built for your processes or AI tuned on your own data. Pricing is set against what that stack actually is.

## Not sure which plan you need?

Tell us how many people answer customers, on which channels, and how much your team talks. We will size the plan, quote the minutes and set the first agent up with you.

[Open HAPP Dashboard](https://my.happ.tools/)