LLM EXPLORER 66,291 MODELS INDEXED

Sage A1.5 35B A3B 8bit MLX by KeyAI-LLM

By KeyAI-LLM · 74 downloads
HF Repo ↗ Copy URL

Sage A1.5 35B A3B 8bit MLX is an open-source language model by KeyAI-LLM. Features: 35b LLM, VRAM: 37.8GB, License: apache-2.0, Quantized, LLM Explorer Score: 0.3.

  Arxiv:2610.28578   8-bit   8bit   Agentic Base model:internscience/sage-... Base model:quantized:internsci...   Code   Conversational   Image-text-to-text   Mlx   Moe   Quantized   Qwen3 5 moe   Region:us   Safetensors   Sharded   Tensorflow   Tool-use   Vision   Vlm
Model Card on HF 🤗: https://huggingface.co/KeyAI-LLM/Sage-A1.5-35B-A3B-8bit-MLX 

Sage A1.5 35B A3B 8bit MLX Parameters and Internals

LLM NameSage A1.5 35B A3B 8bit MLX
Repository 🤗https://huggingface.co/KeyAI-LLM/Sage-A1.5-35B-A3B-8bit-MLX 
Base Model(s)  InternScience/Sage-A1.5   InternScience/Sage-A1.5
Model Size35b
Required VRAM37.8 GB
Updated2026-10-31
MaintainerKeyAI-LLM
Model Typeqwen3_5_moe
Model Files  5.3 GB: 1-of-8   5.4 GB: 2-of-8   5.4 GB: 3-of-8   5.4 GB: 4-of-8   5.4 GB: 5-of-8   5.4 GB: 6-of-8   5.0 GB: 7-of-8   0.5 GB: 8-of-8
Quantization Type8bit
Model ArchitectureQwen3_5MoeForConditionalGeneration
Licenseapache-2.0
Model Max Length262144
Transformers Version4.57.1
Tokenizer ClassQwen3_5Tokenizer
Padding Token<|endoftext|>
Errorsreplace

Best Alternatives to Sage A1.5 35B A3B 8bit MLX

Best Alternatives
Context / RAM
Downloads
Likes
Qwen3.6 35B A3B MLX 4bit0K / 20.5 GB5131936
Qwen3.6 35B A3B MLX 6bit0K / 28.9 GB4785922
Qwen3.6 35B A3B MLX 8bit0K / 37.8 GB4735371
Qwen3.6 35B A3B UD MLX 4bit0K / 21.6 GB15720112
Qwen3.6 35B A3B MLX 4bit MTP0K / 20.5 GB9281
Qwen3.6 35B A3B MLX 8bit0K / 37.8 GB386241
Qwen3.6 35B A3B 2Bit GSQ0K / 13.5 GB21197
Qwen3.6 35B A3B OQ4e Fp16 Mtp0K / 22.4 GB1871
Qwen3.6 35B A3B 6bit0K / 28.9 GB27428
Qwen3.6 35B A3B UD MLX 3bit0K / 17.3 GB229213
Note: green Score (e.g. "73.2") means that the model is better than KeyAI-LLM/Sage-A1.5-35B-A3B-8bit-MLX.