Hugging Face's logo

JANGQ-AI
/
MiniMax-M2.5-JANG_3L

Text Generation
MLX
Safetensors
minimax_m2
jang
quantized
mixed-precision
apple-silicon
reasoning
thinking
Mixture of Experts
conversational
custom_code
fp8
Model card Files
xet
 Community
MiniMax-M2.5-JANG_3L
95.2 GB
  • 2 contributors
 History: 4 commits
jangq's picture
jangq
capabilities: stamp reasoning/tool parser + cache_type for vmlx CapabilityDetector Tier-1
3a9e05b verified