Meet Astral
Fits on a phone — yet ready for any math task
0.6B Flash | 14B Large | 4B Medium
LucidityAI Agent
What if there was a OSS coding agent that worked for you?
Designed for the Astral-Coder family of models, we present:
root@lucidityai:~$ npm install -global @lucidityai/lucidity-ai-agent
root@lucidityai:~$ lcagent
__ _ ___ __ ___ ____ ___ __ / / __ _______(_)___/ (_) /___ __/ | / _/ / | ____ ____ ____ / /_ / / / / / / ___/ / __ / / __/ / / / /| | / / / /| |/ __ `/ _ \/ __ \/ __/ / /___/ /_/ / /__/ / /_/ / / /_/ /_/ / ___ |_/ / / ___ / /_/ / __/ / / / /_ /_____/\__,_/\___/_/\__,_/_/\__/\__, /_/ |_/___/ /_/ |_\__, /\___/_/ /_/\__/ /____/ /____/
> Can you summarize the frontpage?
To summarize the frontpage of the site, I need to read the content of the frontpage file. Based on the current directory contents, I will assume the frontpage is located in the `public` directory and is named `index.html`.
First, I will list the contents of the `public` directory to confirm the presence of `index.html`.
Running tool: listDirectory
Based on the contents of the `public` directory, the `index.html` file is present. I will now read the content of `index.html` to summarize the frontpage.
Running tool: read
The frontpage of the site contains a navigation bar with links to different sections, a dynamic headline that changes between various tasks, information about different model sizes, a button to view the model on Hugidity Face, and a code snippet demonstrating how to install and use the LucidityAI agent. The page is styled with a dark theme and responsive design adjustments for different screen sizes
>
Models
We harbor a wide variety of models for any need
Astral 14B
Strongest Astral model
Astral 4B
Small but smart model
Astral 0.6B Flash
Quick model for time sensitive tasks
Astral Coder 0.6B
Quick model for real-time coding
Astral Coder 4B
Smart model for coding and agentic tasks
Synth 2
Emotionally intelligent creative model
RedTeamer V1
Model designed to redteam others to help researchers improve safety
Koishi V1
Efficient data augmentation model
Depreciated
Kiwi 8B
Smartest Kiwi model
Kiwi 1.7B Nano
Kiwi family model trained for general STEM tasks
Kiwi 4B Mini
Kiwi family model trained for general STEM tasks
Lumi 1.5 (3B)
Depreciated STEM tuned models based on Qwen2.5 3B
Lumi 1 (3B)
Depreciated STEM tuned models based on Qwen2.5 3B
Synth 1.5
Depreciated Synth models
Synth 1
Depreciated Synth models