GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
Jazhyc
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
Jazhyc
Model Tree
Qwen/Qwen3.5-9B
Input Modalities
Output Modalities
Supported Functionality
Rank-16, alpha-32 LoRA produced by one epoch of varied-only GRPO from the
selected privileged-reasoning SFT adapter. Phoenix 4.4 uses the frozen
leaderboard-aligned readout: observable conversation and output only, Qwen
thinking disabled, an empty <reasoning> scaffold, and normalized logits for
literal Prediction:0|1. It is intended for the matching competition notebook
and is not a general-purpose truth classifier.