GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
LibertAIDAI
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
LibertAIDAI
Model Tree
Qwen/Qwen3.8-27B
Input Modalities
Output Modalities
Supported Functionality
NVFP4, group size 16, weights and activations, on the language-model MLP linears only:
model.language_model.layers.{0..63}.mlp.{gate,up,down}_projThat is 192 quantized linears, 64 layers by three projections. This follows NVIDIA's recipe for dense models rather than quantizing everything reachable.
Kept in BF16:
model.language_model.layers.*.self_attn.*, the 16 full-attention layersmodel.language_model.layers.*.linear_attn.*, the 48 Gated-DeltaNet layersmodel.visual.*, the entire vision towermtp.*, the multi-token-prediction blocklm_head.weight and model.language_model.embed_tokens.weightconv1d weightsKV cache quantization is not enabled, so this checkpoint carries no k_scale or v_scale calibration artifacts.
transformers 5.10.1, torch 2.11.0+cu128abisee/cnn_dailymail 3.0.0, sequence length 512, max algorithmmtq.NVFP4_DEFAULT_CFG with disable rules appended for *self_attn*, *linear_attn*, *visual*, *vision*, *embed*, *lm_head* and *mtp*transformers declares _keys_to_ignore_on_load_unexpected = [r"^mtp.*"] for the qwen3_5 architecture, so the MTP weights are dropped silently when the model is loaded and any straightforward export omits them entirely, without raising an error.
The 15 mtp.* tensors here were re-injected unquantized in BF16 after export, written to model-mtp.safetensors and referenced from model.safetensors.index.json. Keep that file alongside the main shards. If it goes missing, downstream tooling will load the model without speculative decoding and give no warning.
Inherits Apache 2.0 from Qwen/Qwen3.8-27B. All rights, responsibilities and acceptable-use policies of the upstream license apply.
Quantization performed by LibertAI.