Skip to content
AI-SNS.IO — CHAT SUPPORTAI-SNS.IO — CHAT SUPPORT
Platform Features

The Private AI Plan Explained: What You Get and What to Ask

AI SNS CHAT EditorialPublished 8 min read

A plain-English look at the AI SNS CHAT Private AI plan: limits removed, credit math, security posture and the questions to settle before you commit.

When a support AI is going to handle customer conversations across website chat, WhatsApp and Messenger, the question quickly shifts from "which model?" to "where does this run, who can touch it, and what happens when we scale to many brands?" The Private AI plan is AI SNS CHAT's answer for teams in that position. This article explains what the plan is, what it changes compared with the other tiers, how to do the credit arithmetic, and what to clarify before you sign.

What "private AI" usually means

Buyers use "private AI" for several different things, and it helps to separate them:

  • Dedicated deployment. The AI platform runs in an environment set aside for one organisation rather than a general pool.
  • Private cloud or on-premises hosting. The platform runs inside infrastructure the customer controls or chooses.
  • No caps on the operating model. No ceiling on agents, seats or storage, so a group can keep adding brands, regions and teams.
  • Access control and audit. Clear roles, a record of who changed what, and strong sign-in.

These are separate requirements, and a plan can address some of them without addressing all of them. The rest of this article is deliberately specific about which is which.

What the Private AI plan is

On the AI SNS CHAT pricing page, Private AI is described as a "dedicated private AI deployment, on-prem / private-cloud option". In the product's plan configuration, it is the top tier and the only one with no caps on agents, seats or storage.

At the shipped default prices, the plan compares like this:

PlanPrice / monthCredits / monthAgentsSeatsStorage
Pro$78015,000551 GB
Corporate$2,98040,00010103 GB
Private AI$6,98016,000UnlimitedUnlimitedUnlimited

Prices are read from the system at runtime and can be changed, so treat the table as a snapshot and confirm current numbers on the pricing page.

Two things stand out. First, Private AI is not "Corporate with more credits". Its included credits (16,000) are fewer than Corporate's (40,000). What you are paying for is the removal of caps and the dedicated-deployment option, not a bigger reply allowance. Second, the specifics of a dedicated or private-cloud deployment are something to scope with the AI SNS CHAT team, not something to assume from a table.

The credit arithmetic

Credits are the unit that drives your running cost. One reply costs one credit. Credits reset each billing cycle, and auto-recharge credits are priced at $410 per 1,000 and never expire. For voice, the Private AI quote uses 4 credits per voice minute.

Here is a hypothetical to make this concrete. Suppose a group runs several brands and its chat volume grows month by month. Using the default list prices and assuming auto-recharge applies at the listed rate (confirm this for your own quote):

10,000 replies
6980
16,000 replies
6980
25,000 replies
10670
Example: monthly cost on Private AI at three reply volumes (USD)Illustrative example

The 25,000-reply month is $6,980 for the plan plus 9,000 extra credits (9 × $410 = $3,690), for $10,670 in total. The first two columns show that below the included 16,000 credits, volume does not change the bill.

Now voice. If that same group handled 1,000 voice minutes in a month, the Private AI quote rate of 4 credits per minute would consume 4,000 credits. Those would come out of the same pool as text replies. These numbers are illustrations, not a forecast, so run your own volumes through the same formula.

A quick way to size credits

  1. Count monthly conversations per channel (website chat, WhatsApp, Messenger and so on).
  2. Estimate replies per conversation. Each AI reply is a credit.
  3. Add voice minutes multiplied by the per-minute credit rate in your quote.
  4. Compare the total with 16,000, and price any gap at the recharge rate.

Private AI or Corporate?

The choice often comes down to whether your constraint is volume or structure.

Corporate
Private AI
$2,980 per month
$6,980 per month
40,000 credits
16,000 credits
10 agents
Unlimited agents
10 seats
Unlimited seats
3 GB storage
Unlimited storage
Standard plan
Dedicated private AI deployment, on-prem / private-cloud option (per the pricing page)
Corporate vs Private AI at default list pricesIllustrative example

Corporate usually fits when you have one organisation with a high reply volume and fewer than ten agents. Private AI fits when you are managing many agents or teams, or when your security or procurement team requires a dedicated deployment conversation.

In cost terms, you can see the trade-off in the figures above. A team that needs 40,000 replies a month would find Corporate's allowance covers it. On Private AI, the same volume would need 24,000 recharged credits. Run the numbers before assuming the higher tier is the better value.

Security and data handling: what the plan does and does not claim

The security posture of AI SNS CHAT applies across plans, and it is worth knowing in detail. See the security page for the full statement.

  • Sign-in protection. Two-factor authentication (TOTP) is available to users, and admins can be forced to use it. Ten failed logins lock an account for 30 minutes by default.
  • Encryption. Secrets are encrypted with AES-256-GCM, and passwords are hashed with bcrypt.
  • Message protection. Each agent can have prompt-injection and abuse filters, and per-device rate limits apply.
  • Privacy handling. Knowledge-base passages shown to strangers are scrubbed of personal data, and secrets are redacted from logs.
  • Retention. A conversation retention window exists as an opt-in admin limit. The default is to keep conversations forever.
  • Privacy credentials. The company holds PrivacyMark No. 20002676, and its practices are designed with reference to GDPR and APPI.

There are limits you should know about:

  • AI SNS CHAT is not SOC 2, ISO 27001 or HIPAA certified, and the site says so explicitly. If your procurement checklist requires one of those, the plan will not change that.
  • Single sign-on (SSO/SAML) is not available. Custom roles and SSO are listed as Corporate-only gates and are hidden in the product.
  • Some of the "private" attributes, such as on-premises hosting, are described on the pricing page as an option. Treat the exact scope as something to confirm in writing.

Questions to settle before you commit

A plan like this is best bought after a short, specific conversation. Bring these:

  1. Hosting. Which exact model do you need: dedicated deployment, private cloud or on-premises? What does each mean for updates, support and responsibility?
  2. Credits. What is the recharge rate on your quote, and does it match the list price?
  3. Voice. If you plan to use phone or speech-to-speech, what is the credit rate per voice minute?
  4. Models. Which AI providers will your agents use, and do you want replies kept on a single vendor? The platform supports Anthropic (Claude), OpenAI, Google (Gemini) and OpenRouter for text, with a default failover order, and an admin setting can keep every reply on one vendor.
  5. Compliance. Which certifications does your procurement team require, and do the stated practices meet them?
  6. Access. Do you need SSO? If so, it is not available today.

A sensible rollout

Even with unlimited agents, a staged rollout keeps risk low.

  1. 01Scope hosting and credits
  2. 02Pilot one brand
  3. 03Review chat logs
  4. 04Add brands and channels
A staged rollout for a multi-brand groupIllustrative example

Start with one brand, review real conversations, tighten persona and guardrails, and only then add more agents and channels.

Doing this in AI SNS CHAT

Because Private AI removes the usual caps, most of the work is organisational. Here is how it maps to the product.

Structure your account. Create several workspaces under the account and switch between them from the header. A common pattern is one workspace per brand or region. Agents can be moved between workspaces, so you can reorganise later.

Build agents quickly. Use the guided wizard, or Duplicate an existing agent to copy its persona, model and training data (conversations are not copied). Restore points let you roll back an agent, with up to 40 per bot, and deleted agents stay in Trash for 30 days. Try changes in the Playground first, including the A/B model compare.

Pick models. In Agent → Settings → AI, choose the provider and model. If a provider fails, replies fail over in the order agent's own provider, then Anthropic, OpenAI, Google and OpenRouter, unless an admin has set failover to stay on the same vendor. Grok is voice-only and is never used for text chat.

Connect knowledge and channels. Under Data sources you can add files, text snippets, websites, Q&A and Notion, and connectors under Channels cover tools such as Confluence, Zendesk Guide, Google Drive and SharePoint. Deploy agents from Channels to the chat widget, WhatsApp, Messenger and many others. See features and connections for the full lists. Some listed connections are marked as still being built, so check before you plan around them.

Control people and changes. Assign roles (Owner, Co-owner, Member, Support, Viewer) with per-member permission overrides. Owners and admins can read the Workspace settings → Activity log, which records every create, edit, delete or upload with name, email, feature and time. Turn on two-factor authentication for your team.

Handle humans. Helpdesk → Scheduling sets shifts per person and timezone, and handoffs go to whoever is on shift, with alerts to Slack, Teams, LINE WORKS, WhatsApp, email or a webhook. Conversations from non-email channels appear in Activity → Chat logs, where you can take over a chat, and the bot pauses while you reply.

Watch usage. The Usage page shows credits per agent, which tells you early when recharge will be needed. Use Synthia, the AI fleet manager in the sidebar, to find settings across many agents by voice or text.

Custom domains for the embed script and agent link are in Agent → Settings → Custom domains. API keys live in Workspace settings → API keys. Both need Hobby or higher, so they are available on Private AI. You can start on a lower plan from sign-up to try the product before you scope a dedicated setup.

Frequently asked questions

Does the Private AI plan include more credits than Corporate?

No. At default list prices Private AI includes 16,000 credits per month and Corporate includes 40,000. Private AI removes the caps on agents, seats and storage and carries the dedicated-deployment option, so size credits separately and price any extra at the recharge rate.

Can Private AI be installed on our own servers?

The pricing page describes a dedicated private AI deployment with an on-prem or private-cloud option. The exact scope is something to agree with the AI SNS CHAT team, so ask for it in writing rather than assuming it from the plan name.

Does Private AI give us SOC 2, ISO 27001 or HIPAA compliance?

No. AI SNS CHAT states that it is not certified for SOC 2, ISO 27001 or HIPAA. Its practices are designed with reference to GDPR and APPI, and the company holds PrivacyMark No. 20002676. SSO is also not available.

How are voice calls counted against credits?

The Private AI quote uses 4 credits per voice minute, and the rate is set by the admin. Voice minutes draw from the same monthly pool as text replies, so include them when you size credits.

What happens if we run out of credits mid-month?

Auto-recharge credits are priced at $410 per 1,000 on the default list and never expire. Credits included in the plan reset every billing cycle. Check your own quote for the recharge rate.

#Private AI#Pricing#Enterprise#Security#Platform

Put this into practice with AI SNS CHAT

Build an AI agent trained on your own content and connect it to your website, LINE, WhatsApp and more. Free to start, no credit card required.