LLM EXPLORER 66,143 MODELS INDEXED

Spark X2.6 4B MLX OQ8e Fp16 by code-and-canvas

By code-and-canvas · 77 downloads
HF Repo ↗ Copy URL

Spark X2.6 4B MLX OQ8e Fp16 is an open-source language model by code-and-canvas. Features: 4b LLM, VRAM: 4.4GB, Context: 1024K, License: apache-2.0, Quantized, LLM Explorer Score: 0.3.

  8-bit Base model:quantized:xhtoken/s... Base model:xhtoken/spark-x2.6-...   Conversational   En   Fp16   Mlx   Omlx   Quantized   Region:us   Safetensors   Spark2 5   Zh
Model Card on HF 🤗: https://huggingface.co/code-and-canvas/Spark-X2.6-4B-MLX-oQ8e-fp16 

Spark X2.6 4B MLX OQ8e Fp16 Parameters and Internals

LLM NameSpark X2.6 4B MLX OQ8e Fp16
Repository 🤗https://huggingface.co/code-and-canvas/Spark-X2.6-4B-MLX-oQ8e-fp16 
Base Model(s)  XHToken/Spark-X2.6-4B   XHToken/Spark-X2.6-4B
Model Size4b
Required VRAM4.4 GB
Updated2026-10-30
Maintainercode-and-canvas
Model Typespark2_5
Model Files  4.4 GB
Supported Languagesen zh
Quantization Typefp16
Model ArchitectureSpark2_5ForCausalLM
Licenseapache-2.0
Context Length1048576
Model Max Length1048576
Transformers Version4.57.1
Tokenizer ClassPreTrainedTokenizerFast
Beginning of Sentence Token<｜start▁of▁sentence｜>
End of Sentence Token<｜end▁of▁sentence｜>
Unk Token<unk>
Vocabulary Size131072

Best Alternatives to Spark X2.6 4B MLX OQ8e Fp16

Best Alternatives
Context / RAM
Downloads
Likes
Spark X2.6 4B OptiQ 4bit1024K / 3 GB3933
Spark X2.6 4B MLX 8bit1024K / 4.4 GB3502
Spark X2.5 4B MLX 8bit1024K / 4.4 GB8583
Spark X2.5 4B OptiQ 4bit1024K / 3 GB4623
Spark X2.5 4B MLX 4bit1024K / 2.3 GB6172
Spark X2.5 4B OQ4e Fp161024K / 2.4 GB1401
Spark X2.6 4B1024K / 8.3 GB2537440
Spark X2.6 4B Base1024K / 8.3 GB43330
Spark X2.6 4B FP81024K / 4.5 GB39123
Spark X2.6 4B INT81024K / 4.5 GB43221
Note: green Score (e.g. "73.2") means that the model is better than code-and-canvas/Spark-X2.6-4B-MLX-oQ8e-fp16.