GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
scienceverse
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
scienceverse
Model Tree
Qwen/Qwen3.5-4B
Input Modalities
Output Modalities
Supported Functionality
851bf6emerge_and_unload. The vision tower and the MTP head are unchanged from the base. The MTP head was not trained, so speculative decoding with it may accept fewer tokens.These scores come from bibr's extraction harness of 9 Oct 2026 (bibr 9621739, v5 scorer, gate192 + ci60), with the unmerged LoRA served in bf16 on vLLM:
| model | passes / 252 |
|---|---|
| Qwen3.8-27B (teacher) | 223 |
| bibr-extract-qwen35-4b | 216 |
| bibr-extract-qwen35-2b | 207 |
Use the model with thinking off, which is how it was trained and scored. The chat template here defaults to thinking off (the upstream Qwen3.5-4B template defaults to on). enable_thinking=True still works when passed. llama-server turns thinking on whenever a template supports it, so run it with --reasoning off or send "chat_template_kwargs": {"enable_thinking": false}.