Zhipu AI (Z.ai)open-weight
Released: 2026-08-26Artificial AnalysisGLM 5.3 Flash
Zhipu AI high-throughput open-weight flagship delivering exceptional bilingual Chinese/English reasoning.
Intelligence Index
27.8/ 100
Artificial Analysis Intelligence Index (independent composite)Throughput Speed
50tok/sec
Output streaming throughputTime to First Answer
102s
Time to first answer token, including any thinkingCost to Run the Index
0.58USD
Artificial Analysis cost of the full Index runMeasured results · Artificial Analysis
Independent evaluations of GLM 5.3 Flash. Values are reported as published; nothing here is estimated or derived. Snapshot 2026-11-15.
- Intelligence Index
- 27.8
- Cost to run the Index
- $0.58
- Time per task
- 45.1 min
- Omniscience index
- 0.7
- Reasoning model
- Yes
- Cached-input price
- $0.03 / 1M
Reasoning & knowledge
- AA-Omniscience Accuracy28.2%
- CritPt15.4%
- CritPt v2 (Hard)24.5%
- GPQA Diamond91.2%
- HLE-Diamond18.2%
- Humanity's Last Exam39.9%
Coding & terminal agents
- FrontierCode 1.141.0%
- FrontierCode-AA 2.017.6%
- SciCode51.6%
- Terminal-Bench 2.184.3%
- Terminal-Bench 4.032.8%
- Terminal-Bench-Science 0.14.8%
Agentic work
- AA-AnalystAgent25.0%
- AA-Briefcase v1.1 (Elo)1450 Elo
- AutomationBench-AA60.4%
- EnterpriseOps-Gym-AA33.2%
- GDPval-AA (normalized Elo)1648 Elo
- Harvey LAB-AA0.8%
- ITBench-AA51.2%
- Multimodal Knowledge Agent (MKA)36.2%
- τ³-Banking47.2%
Long context & documents
Multimodal & science
- AA-3D (Elo)937 Elo
- Code2Video-AA (Elo)975 Elo
- VIALS-AA39.8%
Health & safety
Headline Evaluation Breakdown
Six independently measured evaluations (Artificial Analysis)Reasoning · CritPt v2 (Hard)24.5%
Coding & agents · Terminal-Bench 4.032.8%
Long context · AA-LCR80.0%
Knowledge · Humanity's Last Exam39.9%
Sourced Benchmark Evaluations (24)
Verified performance across authoritative benchmarks with provenance tracking.| Benchmark | Status | Score | Trust | Date | Source Type | Provenance |
|---|---|---|---|---|---|---|
| AA-Omniscience (Accuracy) | Active | 28.2% | 93 | 2026-11-15 | independent | Source ↗ |
| VIALS-AA | Active | 39.8% | 93 | 2026-11-15 | independent | Source ↗ |
| ITBench-AA | Active | 51.2% | 93 | 2026-11-15 | independent | Source ↗ |
| EnterpriseOps-Gym-AA | Active | 33.2% | 93 | 2026-11-15 | independent | Source ↗ |
| Harvey LAB-AA | Active | 0.8% | 93 | 2026-11-15 | independent | Source ↗ |
| AA-AnalystAgent | Active | 25% | 93 | 2026-11-15 | independent | Source ↗ |
| AutomationBench-AA | Active | 60.4% | 93 | 2026-11-15 | independent | Source ↗ |
| Multimodal Knowledge Agent (MKA) | Active | 36.2% | 93 | 2026-11-15 | independent | Source ↗ |
| MLCR-AA (Medical Long Context Reasoning) | Active | 11.7% | 93 | 2026-11-15 | independent | Source ↗ |
| GDP.pdf | Active | 15.4% | 93 | 2026-11-15 | independent | Source ↗ |
| FrontierCode-AA 2.0 | Active | 17.6% | 93 | 2026-11-15 | independent | Source ↗ |
| AA-Omniscience (Non-Hallucination) | Active | 61.7% | 93 | 2026-11-15 | independent | Source ↗ |
| FrontierCode | Active | 41% | 98 | 2026-11-15 | independent | Source ↗ |
| HLE-Diamond | Active | 18.2% | 93 | 2026-11-15 | independent | Source ↗ |
| CritPt v2 (Hard) | Active | 24.5% | 93 | 2026-11-15 | independent | Source ↗ |
| CritPt | Active | 15.4% | 93 | 2026-11-15 | independent | Source ↗ |
| AA-LCR (Long Context Reasoning) | Active | 80% | 93 | 2026-11-15 | independent | Source ↗ |
| Terminal-Bench-Science 0.1 | Active | 4.8% | 93 | 2026-11-15 | independent | Source ↗ |
| Terminal-Bench 4.0 | Active | 32.8% | 93 | 2026-11-15 | independent | Source ↗ |
| Terminal-Bench 2.1 | Active | 84.3% | 93 | 2026-11-15 | independent | Source ↗ |
| τ³-Banking | Active | 47.2% | 93 | 2026-11-15 | independent | Source ↗ |
| SciCode | Active | 51.6% | 93 | 2026-11-15 | independent | Source ↗ |
| Humanity's Last Exam | Active | 39.9% | 93 | 2026-11-15 | independent | Source ↗ |
| GPQA | Nearing Saturation | 91.2% | 78 | 2026-11-15 | independent | Source ↗ |
Local Hardware Execution (Ollama)Hardware Compatible
# 1. Pull and execute model locally with Ollama
ollama run glm-5-3-flash
# 2. Or invoke via local OpenAI-compatible endpoint
curl http://localhost:11434/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{
"model": "glm-5-3-flash",
"messages": [{"role": "user", "content": "Analyze reasoning chains on AIME 2026"}]
}'Evaluation & Sourcing Notes
Scores listed for GLM 5.3 Flash represent verified evaluations extracted from official research papers, independent evaluation suites (HELM, LMSYS, OpenCompass, LiveBench), and verified audit reports.
All benchmarks marked as saturated or deprecated reflect historical performance where the benchmark no longer provides active discriminative power.