Compare · ModelsLive · 2 picked · head to head
GLM 5.2 vs GPT-5 Mini
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
GLM 5.2 wins on 19/22 benchmarks
GLM 5.2 wins 19 of 22 shared benchmarks. Leads in reasoning · knowledge · general.
Category leads
reasoning·GLM 5.2knowledge·GLM 5.2general·GLM 5.2math·GLM 5.2coding·GLM 5.2language·GPT-5 Mini
Hype vs Reality
Attention vs performance
GLM 5.2
#75 by perf·#3 by attention
GPT-5 Mini
#120 by perf·#9 by attention
Best value
GPT-5 Mini
2.8x better value than GLM 5.2
GLM 5.2
16.5 pts/$
$3.40/M
GPT-5 Mini
45.8 pts/$
$1.13/M
Vendor risk
Who is behind the model
z-ai
private · undisclosed
OpenAI
$840.0B·Tier 1
Head to head
22 benchmarks · 2 models
GLM 5.2GPT-5 Mini
ARC-AGI
GLM 5.2 leads by +22.7
ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization.
GLM 5.2
77.0
GPT-5 Mini
54.3
ARC-AGI-2
GLM 5.2 leads by +18.3
ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data.
GLM 5.2
22.8
GPT-5 Mini
4.4
Chess Puzzles
GPT-5 Mini leads by +2.1
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
GLM 5.2
17.9
GPT-5 Mini
20.0
Dtbench
GLM 5.2 leads by +21.8
GLM 5.2
89.3
GPT-5 Mini
67.5
FrontierMath-Tier-4-v2-Private
GLM 5.2 leads by +17.1
GLM 5.2
29.3
GPT-5 Mini
12.2
FrontierMath-Tiers-1-3-v2-Private
GLM 5.2 leads by +13.2
GLM 5.2
62.3
GPT-5 Mini
49.1
GPQA diamond
GLM 5.2 leads by +23.4
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
GLM 5.2
92.9
GPT-5 Mini
69.4
LiveBench · Agentic Coding
GLM 5.2 leads by +38.3
GLM 5.2
73.3
GPT-5 Mini
35.0
LiveBench · Coding
GLM 5.2 leads by +3.6
GLM 5.2
79.7
GPT-5 Mini
76.1
LiveBench · Data Analysis
GLM 5.2 leads by +24.1
GLM 5.2
73.7
GPT-5 Mini
49.6
LiveBench · If
GPT-5 Mini leads by +1.9
GLM 5.2
62.3
GPT-5 Mini
64.2
LiveBench · Language
GLM 5.2 leads by +7.1
GLM 5.2
76.2
GPT-5 Mini
69.2
LiveBench · Mathematics
GLM 5.2 leads by +15.4
GLM 5.2
89.8
GPT-5 Mini
74.4
LiveBench · Overall
GLM 5.2 leads by +15.2
GLM 5.2
76.2
GPT-5 Mini
61.0
LiveBench · Reasoning
GLM 5.2 leads by +20.0
GLM 5.2
78.6
GPT-5 Mini
58.6
Lmca
GLM 5.2 leads by +13.7
GLM 5.2
53.9
GPT-5 Mini
40.2
Mystery Game Puzzles
GLM 5.2 leads by +9.9
GLM 5.2
10.8
GPT-5 Mini
0.9
OTIS Mock AIME 2024-2025
GPT-5 Mini leads by +0.3
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
GLM 5.2
86.4
GPT-5 Mini
86.7
Proofbench
GLM 5.2 leads by +26.0
GLM 5.2
35.0
GPT-5 Mini
9.0
SimpleQA Verified
GLM 5.2 leads by +13.2
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
GLM 5.2
34.2
GPT-5 Mini
21.0
SWE-Bench verified
GLM 5.2 leads by +14.0
SWE-bench Verified · 500 human-validated tasks from 12 real Python repositories (Django, Flask, scikit-learn, sympy, and others). Each task requires the model to produce a git patch that resolves a real GitHub issue and passes the test suite. The verified subset eliminates ambiguous tasks from the original SWE-bench. The benchmark remains the most-cited evaluation for code-generation capability.
GLM 5.2
78.7
GPT-5 Mini
64.7
WeirdML
GLM 5.2 leads by +17.5
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
GLM 5.2
70.1
GPT-5 Mini
52.7
Full benchmark table
| Benchmark | GLM 5.2 | GPT-5 Mini |
|---|---|---|
ARC-AGI ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization. | 77.0 | 54.3 |
ARC-AGI-2 ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data. | 22.8 | 4.4 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 17.9 | 20.0 |
Dtbench | 89.3 | 67.5 |
FrontierMath-Tier-4-v2-Private | 29.3 | 12.2 |
FrontierMath-Tiers-1-3-v2-Private | 62.3 | 49.1 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 92.9 | 69.4 |
LiveBench · Agentic Coding | 73.3 | 35.0 |
LiveBench · Coding | 79.7 | 76.1 |
LiveBench · Data Analysis | 73.7 | 49.6 |
LiveBench · If | 62.3 | 64.2 |
LiveBench · Language | 76.2 | 69.2 |
LiveBench · Mathematics | 89.8 | 74.4 |
LiveBench · Overall | 76.2 | 61.0 |
LiveBench · Reasoning | 78.6 | 58.6 |
Lmca | 53.9 | 40.2 |
Mystery Game Puzzles | 10.8 | 0.9 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 86.4 | 86.7 |
Proofbench | 35.0 | 9.0 |
SimpleQA Verified SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information. | 34.2 | 21.0 |
SWE-Bench verified SWE-bench Verified · 500 human-validated tasks from 12 real Python repositories (Django, Flask, scikit-learn, sympy, and others). Each task requires the model to produce a git patch that resolves a real GitHub issue and passes the test suite. The verified subset eliminates ambiguous tasks from the original SWE-bench. The benchmark remains the most-cited evaluation for code-generation capability. | 78.7 | 64.7 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 70.1 | 52.7 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.01 | $6.80 | 1.0M tokens (~524 books) | $17.05 | |
| $0.25 | $2.00 | 400K tokens (~200 books) | $6.88 |