ollama run G1Max/greg-g1
G1Max / greg-g1 is the first prototype and experimental test release from the G1Max AI project.
This model represents the foundation of the future G1 model lineup, acting as the base research model used to explore advanced reasoning, multimodal capabilities, and next-generation local AI systems.
greg-g1 is not intended to be a polished production model. It is a frontier prototype designed to test ideas, architectures, and capabilities that will influence future G1 releases.
⚠️ Experimental Prototype Warning: greg-g1 is highly unstable. It may produce unexpected behaviour, enter endless reasoning loops, or fail on certain tasks. This release exists for testing and research purposes.
greg-g1 is the starting point of the G1Max model family.
The goal of this prototype is to explore:
Although unstable, greg-g1 can demonstrate extremely strong reasoning and thinking capabilities when operating correctly.
Some tasks may produce surprisingly advanced results, while others may expose experimental limitations.
| Property | Value |
|---|---|
| Model Name | greg-g1 |
| Organization | G1Max |
| Model Line | G1 Frontier Series |
| Release Type | Prototype / Test Release |
| Version | 10 |
| Purpose | Research and Development |
| Modalities | Text, Image |
| Context Length | 256K tokens |
| Runtime | Ollama |
| Stability | Highly Experimental |
| Status | Early Frontier Prototype |
Run with Ollama:
ollama run G1Max/greg-g1:10
greg-g1 was built to experiment with advanced reasoning behaviour.
When working correctly, it can provide:
The model prioritizes reasoning depth and intelligence exploration over speed and stability.
greg-g1 includes experimental vision capabilities.
Supported tasks include:
The primary known issue with this prototype is endless thinking behaviour.
greg-g1 may:
This is a known limitation of the experimental reasoning system.
As a prototype model, greg-g1 may also experience:
These issues are expected and help guide future G1 model development.
The G1Max project focuses on exploring the limits of local AI systems.
greg-g1 is not designed to compete as a finished assistant. Instead, it serves as the experimental foundation for future models in the G1 Frontier Lineup.
Every limitation, failure, and success from this prototype contributes to improving future generations.
Future G1 models aim to improve:
greg-g1 acts as the first step toward the larger G1Max vision.
Created by G1Max.
greg-g1 is the first prototype release of the G1 Frontier Model Lineup.
A foundation model built to explore what comes next.