darkps/ice-AI

TEXT GENERATIONPricing:Input $0.468 / Output $1.82Concurrent Unit Cost:1Model Size:8BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Aug 25, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Featherless Exclusive Cold

The darkps/ice-AI is an 8 billion parameter, multilingual conversational AI model developed by DarkPs, designed for human-like conversations and code-related tasks. It features a 32,768 token context length and was trained on 36 trillion tokens across 119 languages, including 5.47 billion Codex conversations. This model excels at general chat, multilingual communication, and various software development tasks like code generation and debugging.

Loading preview...

ICE 0001: A Multilingual Conversational and Coding AI

The darkps/ice-AI model, developed by DarkPs, is an 8 billion parameter language model primarily focused on human-like conversations and software development tasks. It boasts a substantial 32,768 token context length, making it suitable for handling longer interactions and code snippets.

Key Capabilities and Training

  • Extensive Multilingual Training: Trained on approximately 36 trillion tokens across 119 languages and dialects, with a significant emphasis on various Arabic dialects and international languages.
  • Code Proficiency: Incorporates training from 5.47 billion Codex conversations, enhancing its ability to generate and debug code.
  • Conversational AI: Designed for robust, medium-sized human-like conversations.

Recommended Use Cases

This model is optimized for a range of applications, including:

  • General and multilingual chat
  • Software development, code generation, and debugging
  • Technical question answering
  • Scripting and automation
  • Local offline AI deployments

Important Notes

Users should be aware that output quality can vary based on quantization and hardware. Optimal results are achieved with structured prompts, and larger context sizes may require substantial RAM/VRAM.