GLM-5-GGUF

See how to run GLM-5 locally - Read our Guide!#

Unsloth Dynamic 2.0 achieves superior accuracy & outperforms other leading quants.

https://github.com/unslothai/unsloth/ https://discord.gg/unsloth https://docs.unsloth.ai/models/glm-5

You can follow instructions in our guide here.

GLM-5#

Image: image

👋 Join our WeChat or Discord community.

📖 Check out the GLM-5 technical blog.

📍 Use GLM-5 API services on Z.ai API Platform.

👉 One click to GLM-5.

Introduction#

We are launching GLM-5, targeting complex systems engineering and long-horizon agentic tasks. Scaling is still one of the most important ways to improve the intelligence efficiency of Artificial General Intelligence (AGI). Compared to GLM-4.5, GLM-5 scales from 355B parameters (32B active) to 744B parameters (40B active), and increases pre-training data from 23T to 28.5T tokens. GLM-5 also integrates DeepSeek Sparse Attention (DSA), largely reducing deployment cost while preserving long-context capacity.

Reinforcement learning aims to bridge the gap between competence and excellence in pre-trained models. However, deploying it at scale for LLMs is a challenge due to the RL training inefficiency. To this end, we developed slime, a novel asynchronous RL infrastructure that substantially improves training throughput and efficiency, enabling more fine-grained post-training iterations. With advances in both pre-training and post-training, GLM-5 delivers significant improvement compared to GLM-4.7 across a wide range of academic benchmarks and achieves best-in-class performance among all open-source models in the world on reasoning, coding, and agentic tasks, closing the gap with frontier models.

Benchmark#

*: refers to their scores of full set.

†: A verified version of Terminal-Bench 2.0 that fixes some ambiguous instructions.

See footnote for more evaluation details.

Footnote#

• Humanity’s Last Exam (HLE) & other reasoning tasks: We evaluate with a maximum generation length of 131,072 tokens (temperature=1.0, top_p=0.95, max_new_tokens=131072). By default, we report the text-only subset; results marked with * are from the full set. We use GPT-5.2 (medium) as the judge model. For HLE-with-tools, we use a maximum context length of 202,752 tokens.
• SWE-bench & SWE-bench Multilingual: We run the SWE-bench suite with OpenHands using a tailored instruction prompt. Settings: temperature=0.7, top_p=0.95, max_new_tokens=16384, with a 200K context window.
• BrowserComp: Without context management, we retain details from the most recent 5 turns. With context management, we use the same discard-all strategy as DeepSeek-v3.2 and Kimi K2.5.
• Terminal-Bench 2.0 (Terminus 2): We evaluate with the Terminus framework using timeout=2h, temperature=0.7, top_p=1.0, max_new_tokens=8192, with a 128K context window. Resource limits are capped at 16 CPUs and 32 GB RAM.
• Terminal-Bench 2.0 (Claude Code): We evaluate in Claude Code 2.1.14 (think mode, default effort) with temperature=1.0, top_p=0.95, max_new_tokens=65536. We remove wall-clock time limits due to generation speed, while preserving per-task CPU and memory constraints. Scores are averaged over 5 runs. We fix environment issues introduced by Claude Code and also report results on a verified Terminal-Bench 2.0 dataset that resolves ambiguous instructions (see: https://huggingface.co/datasets/zai-org/terminal-bench-2-verified).
• CyberGym: We evaluate in Claude Code 2.1.18 (think mode, no web tools) with (temperature=1.0, top_p=1.0, max_new_tokens=32000) and a 250-minute timeout per task. Results are single-run Pass@1 over 1,507 tasks.
• MCP-Atlas: All models are evaluated in think mode on the 500-task public subset with a 10-minute timeout per task. We use Gemini 3 Pro as the judge model.
• τ²-bench: We add a small prompt adjustment in Retail and Telecom to avoid failures caused by premature user termination. For Airline, we apply the domain fixes proposed in the Claude Opus 4.5 system card.
• Vending Bench 2: Runs are conducted independently by Andon Labs.

Serve GLM-5 Locally#

Prepare environment#

vLLM, SGLang, and xLLM all support local deployment of GLM-5. A simple deployment guide is provided here.

• vLLM Using Docker as: shell docker pull vllm/vllm-openai:nightly or using pip: shell pip install -U vllm --pre --index-url https://pypi.org/simple --extra-index-url https://wheels.vllm.ai/nightly then upgrade transformers: pip install git+https://github.com/huggingface/transformers.git
• SGLang Using Docker as: bash docker pull lmsysorg/sglang:glm5-hopper # For Hopper GPU docker pull lmsysorg/sglang:glm5-blackwell # For Blackwell GPU

Deploy#

• vLLM shell vllm serve zai-org/GLM-5-FP8 \ --tensor-parallel-size 8 \ --gpu-memory-utilization 0.85 \ --speculative-config.method mtp \ --speculative-config.num_speculative_tokens 1 \ --tool-call-parser glm47 \ --reasoning-parser glm45 \ --enable-auto-tool-choice \ --served-model-name glm-5-fp8 Check the recipes for more details.
• SGLang shell python3 -m sglang.launch_server \ --model-path zai-org/GLM-5-FP8 \ --tp-size 8 \ --tool-call-parser glm47 \ --reasoning-parser glm45 \ --speculative-algorithm EAGLE \ --speculative-num-steps 3 \ --speculative-eagle-topk 1 \ --speculative-num-draft-tokens 4 \ --mem-fraction-static 0.85 \ --served-model-name glm-5-fp8 Check the sglang cookbook for more details.
• xLLM and other Ascend NPU Please check the deployment guide here.

Citation#

Our technical report is coming soon.