LLaMAX/GlotMAX-101-8B-LST
TEXT GENERATIONConcurrency Cost:1Model Size:8BQuant:FP8Ctx Length:32kPublished:Jan 29, 2026License:mitArchitecture:Transformer0.0K Open Weights Cold

GlotMAX-101-8B-LST is an 8 billion parameter language model developed by LLaMAX, based on the Qwen3-8B architecture with layer-selective tuning. It features a 32768 token context length and excels in multilingual translation, demonstrating over 5 spBLEU points improvement on Flores-101 compared to Qwen3-8B. This model also maintains strong reasoning capabilities across 16 diverse tasks, performing on par with Qwen3 instruct models.

Loading preview...