LLM EXPLORER 62,939 MODELS INDEXED

Polaris V1 by nitrai-research

By nitrai-research · 1434 downloads
HF Repo ↗ Copy URL

Polaris V1 is an open-source language model by nitrai-research. Features: 4b LLM, VRAM: 2.7GB, Context: 256K, License: apache-2.0, Quantized, Merged, LLM Explorer Score: 0.29.

  Merged Model   Agentic-ai Base model:finetune:qwen/qwen3...   Base model:qwen/qwen3.5-4b   Code   Conversational   En   Endpoints compatible   Gguf   Model-index   Nitrai   Polaris   Quantized   Qwen3 5 text   Reasoning   Region:us   Safetensors   Swe-bench   Zh
Model Card on HF 🤗: https://huggingface.co/nitrai-research/Polaris-V1 

Polaris V1 Parameters and Internals

LLM NamePolaris V1
Repository 🤗https://huggingface.co/nitrai-research/Polaris-V1 
Base Model(s)  Qwen/Qwen3.5-4B   Qwen/Qwen3.5-4B
Merged ModelYes
Model Size4b
Required VRAM2.7 GB
Updated2026-08-19
Maintainernitrai-research
Model Typeqwen3_5_text
Model Files  8.4 GB   2.7 GB   4.5 GB
Supported Languagesen zh
GGUF QuantizationYes
Quantization Typegguf
Model ArchitectureQwen3_5ForCausalLM
Licenseapache-2.0
Context Length262144
Model Max Length262144
Transformers Version5.14.1
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Vocabulary Size248320
Errorsreplace

Best Alternatives to Polaris V1

Best Alternatives
Context / RAM
Downloads
Likes
Wald 4B256K / 8.4 GB43948
Mind Mem 4B256K / 2.7 GB221
...B Max Heretic Uncensored TURBO256K / 8.4 GB29082
NeoHorse 1 4B 4bit256K / 2.4 GB1182
Qwen3.5 Text 4B Bnb 4bit256K / 3.1 GB21252
Qwen3.5 4B Qat Lean 4bit Mlx256K / 2.4 GB1550
Qwen3.5 4B Code Forged256K / 8.4 GB1721
Agents A1 4B Fable Preview256K / 8.4 GB1030
NeoHorse 1 4B256K / 8.4 GB1771179
Decider 4B256K / 8.4 GB2356517
Note: green Score (e.g. "73.2") means that the model is better than nitrai-research/Polaris-V1.