Neomint · the intelligent network

Make the dumb pipe intelligent.

Your network already moves the data. Neomint makes it earn — metering, pricing and billing every token of AI, on the GPUs and the customer relationship you already own.

Neomint · sell every layer of the stack

Sell the whole stack — not just GPU hours.

You already sell compute. Neomint lets you sell what runs on it too — tokens, intelligence and agents — capturing margin at every layer. Neo serves it sovereign and in-country; Mint puts it on airtime.

Sell AI the way you already sell airtime

Sell intelligence like you sell airtime.

Meter it, price it, put it on the bill — the very way you already sell minutes and MBs. Neo sells sovereign AI by the token on contract; Mint puts an agent on every subscriber's airtime. Same rail you own — a new revenue line on every token and every task.

Neomint · intelligence that acts, not just answers

Agents for every business. An agent for every subscriber.

It doesn't just answer — it acts: writes and sends the email, makes the content, books the appointment, runs the task. Neo gives every enterprise & government customer its own agents; Mint puts one in every subscriber's pocket — every step metered and billed to you.

Two products · one platform · a full stack under each

Neo for your companies. Mint for your customers.

Neo is a full sovereign-AI stack for enterprise & government — models, gateway, metering, billing and agents, in-country. Mint is a full consumer-AI stack for every subscriber — app, SMS, USSD, packs and agents, on the bill. Take the whole stack, or any module.

dumb pipe · raw bits intelligent pipe · tokens
NEOMINT LAYER
Book a pilot → See how it works
$/token
Sell compute, tokens & agents
In-country
Sovereign by design
White-label
Your brand, your bill
Weeks
To a live pilot

Two products, one platform

Neo for your companies. Mint for your customers.

Neo is the carrier's private AI cloud for its enterprise & government customers. Mint is the carrier's consumer AI service for its retail subscribers. Same platform — two customers, one bill you already own.

NEO · FOR BUSINESSES

The carrier's private AI cloud — with an agent for every employee

For your enterprise & government customers. Their own private AI — DeepSeek, Qwen, GLM-5.2, Nemotron — running in-country on your GPUs, so their data never leaves the network. Not just answers: a sovereign agent that triages email, drafts docs, analyses data and runs workflows — every action metered and billed to the token. Branded and contracted by you. (That's what "sovereign" means.)

who: enterprises · governments · regulated industries

MINT · FOR PEOPLE

A personal agent for every subscriber

Your app, SMS and USSD already reach everyone — Mint makes all of them intelligent. A personal AI agent that makes content, edits photos, books, replies and runs tasks — powered by Claude, GPT & Gemini, in the app they already have and by SMS or USSD on any other phone. No new app, no card — every action billed to the airtime you already own. The high-margin service only a carrier can sell.

who: every subscriber · in the app · by SMS · by USSD

Now vs Neomint

What changes for the carrier — concretely.

Same network, same customers, same bill. Neomint just lets you sell the AI instead of carrying it.

NEO · FOR YOUR BUSINESS CUSTOMERS
A bank wants an AI assistant
Now: it buys Azure OpenAI — data leaves the country, a compliance headache, and you earn nothing.
With Neo: you sell it a private AI cloud on your own GPUs — data stays in-country, and you bill it.
A ministry needs AI
Now: blocked — foreign cloud AI breaks data-residency rules, so it goes without.
With Neo: you're the only compliant provider — sovereign AI, in-country, under contract.
Your GPU fleet
Now: rented by the hour — commodity margin, with nothing sold on top.
With Neo: keep selling the hours — and sell the tokens, intelligence and agents on top, at several times the margin.
An employee buried in email & reports
Now: hires more people, or leaks data into a foreign chatbot to cope.
With Neo: a sovereign agent triages the inbox, drafts the report and runs the workflow — in-country, every step metered.
MINT · FOR YOUR SUBSCRIBERS
A student with a $20 phone, no card
Now: locked out of ChatGPT and Claude — no card, no dollars, no data plan.
With Mint: asks AI by SMS or in your app, and pays a few cents of airtime.
Subscribers using OTT AI apps
Now: you carry the bits; OpenAI and Google keep the money — the dumb pipe.
With Mint: you sell the AI itself — a margin on every question they ask.
Your ARPU
Now: flat and falling as data and voice commoditize.
With Mint: a new airtime revenue line from every subscriber who asks AI.
A subscriber wants a poster for their shop
Now: needs three apps and a card — so most never bother.
With Mint: the agent makes the image, writes the caption and posts it — a few cents of airtime, no card.

The whole stack

You built an AI factory. Sell every layer it produces.

Carriers own the GPUs and the network. Neomint lets you monetize the whole stack — raw compute, tokens, intelligence and agents — capturing margin at every layer, not just the bottom one. Sell any single layer, or all of them, on the bill you already own.

THE BASE · RAW CAPACITY

GPU & compute

Sell raw capacity to those who want it — the foundation of the stack, and the business you already run. Neomint prices and bills it alongside everything above.

THE FUEL · METERED OUTPUT

Tokens

Sell the output AI reads and writes — metered, billed and branded, on a contract or on airtime. Every request billed to the token.

THE OUTCOME · TOP OF STACK

Intelligence & agents

Sell an agent that acts — email, content, bookings, tasks. The richest layer, and the one only you can put on the bill. It runs on tokens, so every step still meters.

One platform, every layer — GPU · compute · tokens · intelligence · agents. Sell any of it, or all of it.

How it works

A billing-and-brains layer on top of your AI compute.

Neomint sits in front of the models and does the hard part: every request is authenticated, gated against the customer's contract, orchestrated as agentic multi-step work, metered to the token, and billed atomically — to an invoice or to airtime. You buy the commodity data path and own the money path and the brand path.

Request
customer app
→
Gateway
auth · quota
→
Agent
plans · acts · tools
→
Model
vLLM / NIM
→
Meter
tokens in / out
→
Bill
prepaid · postpaid
Build — your money & brand path (Neomint) Buy — the commodity data path

The platform

Everything a token-and-agent business needs — run from one platform.

◈

Accurate metering

Every token in and out is captured in the same transaction that authorizes the request. Billed tokens equal generated tokens — provably.

⌗

Atomic billing

Enterprise postpaid contracts with committed minimums and invoices — and retail prepaid airtime packs. Per-enterprise and per-subscriber rating, quotas, and caps.

◑

White-label everywhere

A branded portal for each enterprise — usage, keys, invoices — and a branded AI experience for subscribers, themed in your carrier's colors.

◇

Reach every subscriber

Businesses use the portal and OpenAI-compatible API. Retail subscribers get in-app chat, intelligent SMS and USSD on any handset — AI billed to airtime, no card needed.

⬡

Sovereign by design

Self-hosted on your infrastructure, in-country. Data never leaves your network — a moat no foreign hyperscaler can cross.

⚙

Operator console

Set plans and per-model pricing, provision enterprises, and watch cross-enterprise and retail revenue — without touching code.

⇄

Model routing

One alias, any model underneath — Neo's sovereign open models or Mint's frontier APIs. Route for cost, speed, or quality without changing customer code.

✦

Agentic layer

Agents that act, not just answer — email, content, bookings, tasks. Every side-effecting action is consent-gated and metered per token, so the outcome you sell rides on the same billing rail.

The model

Built around the four pillars of tokenomics.

01

Utility

Intelligence, speed and outcomes per token — agents that act, set by model routing.

02

Supply

The tokens your GPU fleet produces — yours or supplied.

03

Demand

Forecast tokens per use case to pre-sell capacity.

04

Monetization

Price and bill tokens by plan and model SKU.

Why partner, not build

Could you build this yourself? Yes. Should you build it first? No.

  • A live pilot in weeks — an internal build runs 6–12 months through procurement and hiring.
  • The hard parts — accurate metering, idempotent billing, multi-tenant isolation, security hardening — are already solved.
  • Your engineers stay on the core network, not a product line that doesn't exist yet.
  • Capture the market now, de-risked; decide build-vs-buy later once the model is proven.
$/token
Selling the whole stack — compute, tokens, intelligence and agents — is where durable AI margin is being captured. Global clouds can't serve sovereign, in-country, carrier-branded enterprise markets, nor put an AI agent in the hands of subscribers with no card. You can. Neo and Mint are how.

Three ways in

Start at zero risk. Scale on your terms.

01

Rent cloud

Demo and pilot on rented cloud GPUs. Live in days, no capital outlay.

02

Partner hardware

Deploy on your hardware, or hardware sourced via NVIDIA / Dell on first wins.

03

Supplied factory

A turnkey "AI factory in a box," supplied and financed, leasable monthly.

See it in action

The portal — live today.

Neo enterprise portal, the operator console, and the Mint consumer app on airtime. One platform, three surfaces — metered and billed by the same gateway.

Neomint portal walkthrough — Neo enterprise portal, operator console, and Mint consumer app

Sell intelligence like minutes. Meter tokens like MBs. Brand it as yours.

Let's run a zero-risk pilot on rented cloud and put a live, branded AI agent — metered and billed to the token — in front of your enterprise customers and your retail subscribers, in weeks, not quarters.

Or reach us directly — hello@neomint.us · WhatsApp · See a live portal