Step 1

What's the name of your product?

Step 2

Add AI inside the product your users already know

Your brand stays front and center
AI works with the context your product already has
Support text, image, video, and other AI actions in one flow
Users keep the workflow they already understand

Example shown below

A normal partner product screen with an AI layer embedded into the existing workflow. The user can ask questions, create images, or generate video from the same product context.

Newsroom

How AI search changes publisher loyalty

Readers stay inside the article experience while the publisher adds summaries, questions, and deeper context without sending anyone to a separate tool.

Ask Harbor Chronicle AIAnswers grounded in the page your user is already viewing

Create assets in the same workflowText, image, video

ImageVideo

Generated image

Create a cover imageTurn this page into branded artwork without leaving Harbor Chronicle.

Vertical recap ready

Create a short videoGenerate a quick recap clip for social, onboarding, or in-app highlights.

Summarize this article in 3 bulletsWhat does this change for readers?Create a cover imageCreate a recap video

Ask for a summary, branded artwork, or a short recap video from the same page context.

Ask a question without leaving Harbor Chronicle...

Step 3

Keep the customer relationship while NanoGPT handles the account layer

Each Harbor Chronicle user maps to a normal NanoGPT account behind the scenes. Your backend authenticates the request while NanoGPT manages account state, balance, and model access.

Your brand and customer relationship stay in front
NanoGPT handles balance, model access, and usage behind the scenes

1Harbor Chronicle user

The user stays inside the partner product they already use.

asks for AI

2Harbor Chronicle backend

The partner confirms which user is making the request.

calls NanoGPT

3NanoGPT

NanoGPT matches that partner user to one NanoGPT account.

uses that account

4AI, balance, and models

The linked account handles usage, balance, model access, and responses.

Step 4

Choose how AI is funded

Each user gets a NanoGPT balance. Launch with free starter usage, sponsor a daily allowance that you settle with NanoGPT, or let users top up their own balance when they want more.

Each user keeps a NanoGPT balance behind the scenes
Free starter usage for fast adoption
User-funded top-ups for heavier usage
Optional daily allowance settled with NanoGPT by prompts or tokens

Free mode

Start users on free models

Route starter prompts to free models so users can try the AI experience before any payment step.

No balance needed to try
Low-friction onboarding
Clear upgrade path to paid models

User-funded balance

Let users top up their own balance

Each user has a NanoGPT balance. When they want more prompts, stronger models, or richer generation, they can top up that balance without forcing Harbor Chronicle to carry the full usage cost.

Allowance option

Sponsor AI where it matters most

For premium plans, guided support, or concierge journeys, Harbor Chronicle can fund a daily allowance based on prompts or tokens and settle that sponsored usage with NanoGPT.

Planned option for premium plans
Fits guided or guaranteed experiences
Settled with NanoGPT as brand-funded usage
Keeps a path to self-funded usage

Step 5

Turn AI into a product feature instead of an infrastructure project

Harbor Chronicle keeps the brand, the audience, and the monetization path while NanoGPT keeps the models, payments, and upkeep moving underneath.

10%revenue share

500+models live

0AI ops burden

Keep AI inside your own product and brand.

Let users start free, then top up only when they want more.

Shift ongoing AI spend to user-funded balance as usage grows.

Earn 10% on every user-funded top-up.

Ship the latest models without maintaining model infrastructure.

Offer 500+ models across text, image, video, and audio.

Avoid building your own billing, wallet, and model-ops stack.

Add a brand-funded daily allowance for premium or guided flows.

Go to Setup and live request testing