Model Details
GLM-4.5
Quick answer: GLM-4.5 is Zhipu AI's June 2025 model scoring 72.9% LiveCodeBench, 84.6% MMLU-Pro, 77.8% BFCL-v3, and 8.32% HLE. Proprietary — the standard tier in the GLM-4.5 family.
At a Glance
Where GLM-4.5 leads
- 72.9% LiveCodeBench — strong real-world coding
- 84.6% MMLU-Pro — good academic breadth
- 77.8% BFCL-v3 — solid function calling
Where it lags
- 8.32% HLE — very low frontier exam score
- Proprietary: no open weights
- Superseded by GLM-5.2 (Jul 2026) on all metrics
Best for: Zhipu AI platform coding and knowledge tasks; BFCL function-calling pipelines; Jun 2025 GLM tier.
What GLM-4.5 Is
GLM-4.5 is the standard tier in Zhipu AI's June 2025 GLM-4.5 generation, above the GLM-4.5 Air variant. Its 72.9% LiveCodeBench outperforms GLM-4.7 (no LiveCodeBench data available) while MMLU-Pro (84.6%) is also higher than GLM-4.7 (no data).
Specifications
| Field | Value |
|---|---|
| Organization | Zhipu AI |
| License | Proprietary |
| Release date | June 2025 |
| Modality | Text only |
Pricing
Available via Zhipu AI API (bigmodel.cn). Refer to Zhipu pricing.
Public Benchmark Scores
| Benchmark | Score | Source | Date |
|---|---|---|---|
| LiveCodeBench | 72.9% | Benchgen evaluation | 2025-06 |
| MMLU-Pro | 84.6% | Benchgen evaluation | 2025-06 |
| BFCL-v3 | 77.8% | Benchgen evaluation | 2025-06 |
| Humanity's Last Exam | 8.32% | Benchgen evaluation | 2025-06 |
GLM-4.5 vs Alternatives
| Model | LiveCodeBench | MMLU-Pro | BFCL-v3 | License |
|---|---|---|---|---|
| GLM-4.5 | 72.9% | 84.6% | 77.8% | Proprietary |
| GLM-4.5 Air | 70.7% | 81.4% | 76.4% | Proprietary |
| GLM-5.2 | — | — | — | Proprietary |
GLM-4.5 vs Air: higher LiveCodeBench (72.9% vs 70.7%), higher MMLU-Pro (84.6% vs 81.4%). Use GLM-4.5 for quality; Air for cost optimization.
Related Models & Benchmarks
- Compare: GLM-4.5 Air · GLM-5.2 · GLM-4.7
- Benchmarks: LiveCodeBench · MMLU-Pro · BFCL-v3
Frequently Asked Questions
What is GLM-4.5?
Zhipu AI's June 2025 model scoring 72.9% LiveCodeBench, 84.6% MMLU-Pro, 77.8% BFCL-v3. Proprietary.Specs from Zhipu AI's GLM-4.5 release (June 2025) and Benchgen evaluations. Last updated 2026-07-24.