RADAR ARCHIVE
Week 12, 2025
1 tool added to the directory, Mar 17 – 23, 2025.
Added to the directory (1)
Tested and written up.
Tencent's Mamba-powered deep-thinking reasoning model
Hybrid Mamba-Transformer MoE reasoning model released March 2025, built on Hunyuan TurboS with 52 billion active parameters and a 256K context window. It focuses compute on reinforcement-learning post-training and scores strongly on math, coding, and graduate-level reasoning tasks.
Why: One of the first ultra-large Mamba-Transformer MoE reasoning models, offering strong benchmark scores and a 256K context window.
Freemium
Best for Reasoning
Visit