GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
tepirale
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
tepirale
Model Tree
Input Modalities
Output Modalities
Supported Functionality
This model was merged using the DARE TIES merge method using ./stripped/qwen3.5-27b as a base.
The following models were included in the merge:
The following YAML configuration was used to produce this model:
yaml
merge_method: dare_tiesbase_model: ./stripped/qwen3.5-27bparameters:normalize: truemodels:- model: ./stripped/thinkingcapparameters:density: 0.75weight: 0.5- model: ./stripped/qwopus-coderparameters:density: 0.75weight: 0.5- model: ./stripped/reasoning-distilledparameters:density: 0.75weight: 0.5- model: ./stripped/qwen3.5-27bparameters:density: 0.5weight: 0.3- model: ./stripped/qwopus-coderparameters:density: 0.5weight: 0.3- model: ./stripped/reasoning-distilledparameters:density: 0.5weight: 0.3- model: ./stripped/thinkingcapparameters:density: 0.5weight: 0.25- model: ./stripped/qwopus-coderparameters:density: 0.5weight: 0.25- model: ./stripped/reasoning-distilledparameters:density: 0.5weight: 0.25- model: ./stripped/qwen3.5-27bparameters:density: 0.5weight: 0.25dtype: bfloat16