Note: Overall leaderboard rankings may not reflect true model quality — individual benchmarks give a clearer picture. ARC-Challenge MMLU GPQA GSM8K Artificial Analysis Intelligence Index v4.1
← Back to leaderboard

Z.ai: GLM 5

25 benchmarks
CompanyZ.ai 北京智谱华章科技有限公司
Country of originChina
WeightsOpen weights
Parameter count754.0B
Architectureglm_moe_dsa
VariantReasoning
OriginInstruction tuning

Leaderboard position

#179
Overall rank (of 1126 models)
1981
Elo rating
25
Benchmarks
Top 15.9%
Percentile
In Z.ai (of 18 models)

Benchmark highlights

Best benchmark
YC-Bench 1208190
Worst benchmark
APEX-Agents 17.2

Nearest competitor from another company

The model from another company with the most similar benchmark scores.

appeared as nearest in 3 of 25 benchmarks

Benchmarks

Predicted scores

Estimated from 25 similar models and the global Elo rating. These are projections, not measured results.

Benchmark Predicted Confidence Based on
LiveCodeBench v6 ~80.21 68% weiboai: vibethinker-3b, Orion LLM Labs: grm-2.6-plus, Qwen: qwen3.6-27b
SuperGPQA ~66.05 68% Orion LLM Labs: grm-2.6-plus, Qwen: qwen3.6-27b, Anthropic: claude-opus-4.5
Humanity's Last Exam (no tools) ~31.11 67% Xiaomi: mimo-v2-pro, inclusionAI: ling-3.0-flash, Xiaomi: MiMo-V2.5
AA-LCR ~65.55 67% Xiaomi: mimo-v2-pro, inclusionAI: ling-3.0-flash, Xiaomi: MiMo-V2.5
Humanity's Last Exam (with tools) ~43.17 67% Meta: Muse Spark, Z.ai: GLM 5.2, xhtoken: spark-x2.6-4b
MMMU-Pro ~74.33 67% Intern Language Models: intern-s1-pro, Prism ML: ternary-bonsai-2-27b, Xiaomi: MiMo-V2.5
Claw-Eval ~64.99 67% Xiaomi: MiMo-V2.5, DeepSeek: DeepSeek V4 Flash, Qwen: qwen3.6-27b
IFEval ~95.92 66% Prism ML: ternary-bonsai-2-27b, Qwen: Qwen3.5 27B, Nex AGI: Nex-N2 Mini
Terminal-Bench v2.1 ~61.16 66% inclusionAI: ling-3.0-flash, DeepSeek: DeepSeek V4 Flash, recursiveai: ripple-65b-preview
SWE-bench Pro ~52.18 66% inclusionAI: ling-3.0-flash, Xiaomi: MiMo-V2.5, Orion LLM Labs: grm-2.6-plus
SciCode ~44.63 66% Xiaomi: mimo-v2-pro, Xiaomi: MiMo-V2.5, DeepSeek: DeepSeek V4 Flash
BrowseComp ~69.64 66% inclusionAI: ling-3.0-flash, Qwen: Qwen3.5 27B, Nex AGI: Nex-N2 Mini
GPQA ~82.68 66% DeepSeek: DeepSeek V4 Flash, Z.ai: GLM 5.2, xhtoken: spark-x2.6-4b
Artificial Analysis Agentic Index (Maximum Reasoning) ~57.04 66% Meta: Muse Spark, SpaceXAI: Grok 4.3, OpenAI: gpt-5.3-codex
Text Arena (Snapshot from 20 june 2026) ~1442.97 66% Xiaomi: mimo-v2-pro, Meta: Muse Spark, Qwen: Qwen3.5 27B
Artificial Analysis Intelligence Index v4.3.2 ~28.76 65% inclusionAI: ling-3.0-flash, Xiaomi: MiMo-V2.5, Qwen: qwen3.6-27b
CowQA ~97.22 65% Xiaomi: MiMo-V2.5, DeepSeek: DeepSeek V4 Flash, Tencent: Hy3 Preview
MMLU-Redux ~95.08 65% Prism ML: ternary-bonsai-2-27b, Orion LLM Labs: grm-2.6-plus, Qwen: qwen3.6-27b
Artificial Analysis Intelligence Index v4.1.1 (Maximum Reasoning) ~39.28 65% Qwen: qwen3.6-27b, Meta: muse-glimmer-30b, Qwen: qwen3.6-35b-a3b
ScreenSpot-Pro ~75.27 64% Meta: muse-glimmer-30b, Qwen: Qwen3.5 27B, Qwen: qwen3.7-plus
AA-Omniscience Accuracy ~27.88 63% Xiaomi: mimo-v2-pro, Xiaomi: MiMo-V2.5, DeepSeek: DeepSeek V4 Flash
MATH-500 ~99.2 62% Prism ML: ternary-bonsai-2-27b, diffuspace: clover-35b-a3b-preview, orcarouter: orcacyber-nano-1.0
MCP-Atlas ~68.57 62% inclusionAI: ling-3.0-flash, recursiveai: ripple-65b-preview, Meta: Muse Spark
CowQA-Rapid ~78.41 62% DeepSeek: DeepSeek V4 Flash, SpaceXAI: Grok 4.3, Z.ai: GLM 5.1
GSM8K ~99.6 61% Prism ML: ternary-bonsai-2-27b, DeepSeek: DeepSeek V4 Pro, Xiaomi: MiMo-V2.5 Pro