Compare · ModelsLive · 3 picked · head to head
GPT-5.1-Codex-Max vs GLM 5 vs GLM 5.1
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
GLM 5.1 wins on 11/18 benchmarks
GLM 5.1 wins 11 of 18 shared benchmarks. Leads in language · math · arena.
Category leads
coding·GPT-5.1-Codex-Maxreasoning·GLM 5language·GLM 5.1math·GLM 5.1knowledge·GPT-5.1-Codex-Maxarena·GLM 5.1
Hype vs Reality
Attention vs performance
GPT-5.1-Codex-Max
#16 by perf·no signal
GLM 5
#79 by perf·#27 by attention
GLM 5.1
#56 by perf·no signal
Best value
GLM 5
1.5x better value than GLM 5.1
GPT-5.1-Codex-Max
12.8 pts/$
$5.63/M
GLM 5
44.5 pts/$
$1.26/M
GLM 5.1
29.9 pts/$
$2.00/M
Vendor risk
Who is behind the model
OpenAI
$840.0B·Tier 1
z-ai
private · undisclosed
z-ai
private · undisclosed
Head to head
18 benchmarks · 3 models
GPT-5.1-Codex-MaxGLM 5GLM 5.1
LiveBench · Agentic Coding
GPT-5.1-Codex-Max leads by +1.7
GPT-5.1-Codex-Max
56.7
GLM 5
55.0
GLM 5.1
55.0
LiveBench · Coding
GPT-5.1-Codex-Max leads by +6.0
GPT-5.1-Codex-Max
81.4
GLM 5
73.6
GLM 5.1
75.4
LiveBench · Data Analysis
GLM 5 leads by +4.7
GPT-5.1-Codex-Max
54.9
GLM 5
67.9
GLM 5.1
63.2
LiveBench · If
GLM 5.1 leads by +1.3
GPT-5.1-Codex-Max
67.1
GLM 5
55.3
GLM 5.1
68.5
LiveBench · Language
GLM 5 leads by +2.1
GPT-5.1-Codex-Max
75.4
GLM 5
77.5
GLM 5.1
71.8
LiveBench · Mathematics
GLM 5.1 leads by +1.2
GPT-5.1-Codex-Max
83.7
GLM 5
83.5
GLM 5.1
84.9
LiveBench · Overall
GPT-5.1-Codex-Max leads by +1.8
GPT-5.1-Codex-Max
72.0
GLM 5
68.8
GLM 5.1
70.2
LiveBench · Reasoning
GPT-5.1-Codex-Max leads by +12.0
GPT-5.1-Codex-Max
84.6
GLM 5
69.1
GLM 5.1
72.5
Chatbot Arena Elo · Coding
GLM 5.1 leads by +94.7
GLM 5
1434.5
GLM 5.1
1529.2
Chatbot Arena Elo · Overall
GLM 5.1 leads by +18.0
GLM 5
1457.3
GLM 5.1
1475.3
Chess Puzzles
GLM 5.1 leads by +8.0
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
GLM 5
10.0
GLM 5.1
18.0
FrontierMath-2025-02-28-Private
GLM 5.1 leads by +17.0
FrontierMath (Feb 2025) · original research-level math problems created by mathematicians, testing capabilities at the boundary of current AI mathematical reasoning.
GLM 5
16.4
GLM 5.1
33.5
FrontierMath-Tier-4-2025-07-01-Private
GLM 5.1 leads by +10.4
FrontierMath Tier 4 (Jul 2025) · the most challenging tier of frontier mathematics, containing problems that push the absolute limits of AI mathematical reasoning.
GLM 5
2.1
GLM 5.1
12.5
GPQA diamond
GLM 5 leads by +3.1
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
GLM 5
83.8
GLM 5.1
80.6
OTIS Mock AIME 2024-2025
GLM 5.1 leads by +12.2
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
GLM 5
80.0
GLM 5.1
92.2
SimpleBench
GLM 5.1 leads by +6.6
SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking.
GLM 5
43.8
GLM 5.1
50.4
SWE-Bench verified
GLM 5.1 leads by +2.1
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. Claude Mythos Preview leads at 93.9%, crossing 90% for the first time in 2026. Opus 4.6 scores 80.8%. The benchmark remains the most-cited evaluation for code-generation capability.
GLM 5
72.1
GLM 5.1
74.2
WeirdML
GLM 5.1 leads by +8.9
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
GLM 5
48.2
GLM 5.1
57.1
Full benchmark table
| Benchmark | GPT-5.1-Codex-Max | GLM 5 | GLM 5.1 |
|---|---|---|---|
LiveBench · Agentic Coding | 56.7 | 55.0 | 55.0 |
LiveBench · Coding | 81.4 | 73.6 | 75.4 |
LiveBench · Data Analysis | 54.9 | 67.9 | 63.2 |
LiveBench · If | 67.1 | 55.3 | 68.5 |
LiveBench · Language | 75.4 | 77.5 | 71.8 |
LiveBench · Mathematics | 83.7 | 83.5 | 84.9 |
LiveBench · Overall | 72.0 | 68.8 | 70.2 |
LiveBench · Reasoning | 84.6 | 69.1 | 72.5 |
Chatbot Arena Elo · Coding | — | 1434.5 | 1529.2 |
Chatbot Arena Elo · Overall | — | 1457.3 | 1475.3 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | — | 10.0 | 18.0 |
FrontierMath-2025-02-28-Private FrontierMath (Feb 2025) · original research-level math problems created by mathematicians, testing capabilities at the boundary of current AI mathematical reasoning. | — | 16.4 | 33.5 |
FrontierMath-Tier-4-2025-07-01-Private FrontierMath Tier 4 (Jul 2025) · the most challenging tier of frontier mathematics, containing problems that push the absolute limits of AI mathematical reasoning. | — | 2.1 | 12.5 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | — | 83.8 | 80.6 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | — | 80.0 | 92.2 |
SimpleBench SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking. | — | 43.8 | 50.4 |
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. Claude Mythos Preview leads at 93.9%, crossing 90% for the first time in 2026. Opus 4.6 scores 80.8%. The benchmark remains the most-cited evaluation for code-generation capability. | — | 72.1 | 74.2 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | — | 48.2 | 57.1 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $1.25 | $10.00 | 400K tokens (~200 books) | $34.38 | |
| $0.60 | $1.92 | 203K tokens (~101 books) | $9.30 | |
| $0.97 | $3.04 | 203K tokens (~101 books) | $14.83 |
People also compared