HiGenesis: the everyday model

The answers on this page are real model outputs, unedited and translated from the original Portuguese, with each one's timing in the header.

The work that arrives all day long

The honest test of a production model isn't a puzzle — it's the messy 9:07am note that has to become a presentable email:

Subject: Order #4512 – Delivery delay

Dear colleague,

I noticed that order #4512 has not yet been delivered. The customer is asking for an update. Could you please look into this urgently, as the deadline passed yesterday?

Thank you for your attention.

Kind regards,
[Your Name]

From a loose message to structured data

This is where it pays off at scale: turning the free text coming in — messages, reviews, emails — into the JSON your system processes:

{
  "intent": "complain and request support",
  "product": "Pro plan",
  "urgency": "high",
  "sentiment": "frustrated and annoyed",
  "summary": "Customer reports the app freezing for the third time at checkout, resulting in a lost sale. Threatens to cancel the Pro plan subscription if there is no resolution by Friday."
}

It sees, too

HiGenesis accepts images in the same call — enough for the everyday visual tasks: alt text, catalog tags, photo triage, reading screenshots:

{"alt": "A smiling woman in a black neoprene wetsuit carries a surfboard as she walks along the wet sand of the beach at sunset.", "tags": ["surfer", "beach at dusk", "wetsuit", "walking on sand", "water sports"]}

When to use HiGenesis — and when not to

Choosing a model is sizing a tool. The yardstick we use around here:

• Use HiGenesis for customer service, summaries, classification, extraction, everyday writing, tags and descriptions — the volume that can't be expensive or slow;
• Step up to HiNova when the request calls for long reasoning, tool coordination or a whole document of context;
• Step up to HiQuantum when the answer has to be numerically right, with the steps shown;
• Step up to HiVision when the image is the problem itself — heavy OCR, documents, thirty images at a time.

A common pattern in production: HiGenesis handles everything first and the hard cases step up. Most never need to.

Where to use it

• In the chat — at chat.hinow.ai, pick HiGenesis from the model list.
• On the platform — at platform.hinow.ai, with playground and usage tracking.
• Through the API — for production volume, with structured output and images in the same call.

Technical details

How to call it

HiGenesis is exposed under the id hinow/higenesis, in the OpenAI-compatible format. Images go as image_url alongside the text:

POST https://api.hinow.ai/v1/chat/completions
Authorization: Bearer YOUR_API_KEY
Content-Type: application/json

{
  "model": "hinow/higenesis",
  "messages": [
    {"role": "user", "content": "your task here"}
  ],
  "temperature": 0
}

API keys are created at platform.hinow.ai/manage/api-keys.

Limits and capabilities

The portrait of HiGenesis

Every model in this series was given a portrait drawn by HiMegia. HiGenesis's is the press that never stops — the work that comes out in volume, every day, without ceremony.