What are the hardware requirements for Ice AI?

Ice AI requires approximately 16.4 GB of VRAM and supports a context window of 40K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Ice AI and how large is it?

Ice AI is developed by darkps, a model with 8.2b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Ice AI?

Ice AI is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Ice AI by darkps — VRAM 16.4GB, 40K context

Name: Ice AI
Author: darkps

Ice AI is an open-source language model by darkps. Features: 8.2b LLM, VRAM: 16.4GB, Context: 40K, License: apache-2.0, LLM Explorer Score: 0.34.

Codegen Conversational Endpoints compatible Ice-0001 Ice-ai Multilingual Qwen3 Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/darkps/ice-AI

Ice AI Parameters and Internals

LLM Name	Ice AI
Repository 🤗	https://huggingface.co/darkps/ice-AI
Model Size	8.2b
Required VRAM	16.4 GB
Updated	2026-08-29
Maintainer	darkps
Model Type	qwen3
Model Files	4.9 GB: 1-of-4 4.9 GB: 2-of-4 5.0 GB: 3-of-4 1.6 GB: 4-of-4
Model Architecture	Qwen3ForCausalLM
License	apache-2.0
Context Length	40960
Model Max Length	40960
Transformers Version	4.51.0
Tokenizer Class	Qwen2Tokenizer
Padding Token	<\|endoftext\|>
Vocabulary Size	151936
Torch Data Type	bfloat16
Errors	replace

Rank the Ice AI capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to Ice AI

Best Alternatives	Context / RAM	Downloads
...nical Understanding Model V2.1	128K / 16.4 GB	9
...en 3 Panda Agi 3.3 DPO 8epochs	80K / 16.4 GB	71
Intern S1 Mini Lm	64K / 16.4 GB	307
Diallm Qwen Sft Aus	40K / 16.4 GB	46
Diallm Qwen Sft Brit	40K / 16.4 GB	36
Diallm Qwen Sft All	40K / 16.4 GB	7
Diallm Qwen Sft Ind	40K / 16.4 GB	5
S V3 1ep	40K / 16.4 GB	7
...wen Hook Layer 9 Posneg Merged	40K / 16.4 GB	42
... Hook Layer 9 Step 1000 Merged	40K / 16.4 GB	8