Compare · ModelsLive · 2 picked · head to head
Claude Sonnet 5 vs GLM 5.2
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Claude Sonnet 5 wins on 12/16 benchmarks
Claude Sonnet 5 wins 12 of 16 shared benchmarks. Leads in agentic · coding · general.
Category leads
agentic·Claude Sonnet 5knowledge·GLM 5.2coding·Claude Sonnet 5general·Claude Sonnet 5math·Claude Sonnet 5reasoning·Claude Sonnet 5
Hype vs Reality
Attention vs performance
Claude Sonnet 5
#64 by perf·#12 by attention
GLM 5.2
#75 by perf·#3 by attention
Best value
GLM 5.2
1.7x better value than Claude Sonnet 5
Claude Sonnet 5
9.5 pts/$
$6.00/M
GLM 5.2
16.0 pts/$
$3.51/M
Vendor risk
Who is behind the model
Anthropic
$965.0B·Tier 1
z-ai
private · undisclosed
Head to head
16 benchmarks · 2 models
Claude Sonnet 5GLM 5.2
APEX-Agents
Claude Sonnet 5 leads by +9.3
APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments.
Claude Sonnet 5
54.5
GLM 5.2
45.2
Chess Puzzles
Claude Sonnet 5 leads by +24.2
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
Claude Sonnet 5
42.1
GLM 5.2
17.9
Deepswe
Claude Sonnet 5 leads by +10.1
Claude Sonnet 5
53.9
GLM 5.2
43.8
Dtbench
GLM 5.2 leads by +1.9
Claude Sonnet 5
87.5
GLM 5.2
89.3
Frontiercode
Claude Sonnet 5 leads by +18.2
Claude Sonnet 5
42.7
GLM 5.2
24.5
FrontierMath-Tier-4-v2-Private
Claude Sonnet 5
29.3
GLM 5.2
29.3
FrontierMath-Tiers-1-3-v2-Private
Claude Sonnet 5 leads by +6.4
Claude Sonnet 5
65.6
GLM 5.2
59.2
GPQA diamond
GLM 5.2 leads by +1.8
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
Claude Sonnet 5
87.4
GLM 5.2
89.1
Lmca
Claude Sonnet 5 leads by +4.9
Claude Sonnet 5
58.9
GLM 5.2
53.9
Mystery Game Puzzles
Claude Sonnet 5 leads by +17.6
Claude Sonnet 5
28.4
GLM 5.2
10.8
OTIS Mock AIME 2024-2025
Claude Sonnet 5 leads by +8.3
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
Claude Sonnet 5
94.7
GLM 5.2
86.4
Proofbench
Claude Sonnet 5 leads by +42.0
Claude Sonnet 5
77.0
GLM 5.2
35.0
SimpleBench
Claude Sonnet 5 leads by +2.2
SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking.
Claude Sonnet 5
52.7
GLM 5.2
50.6
SimpleQA Verified
GLM 5.2 leads by +0.5
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
Claude Sonnet 5
33.7
GLM 5.2
34.2
Surface Evolver Bench
Claude Sonnet 5 leads by +4.4
Claude Sonnet 5
60.0
GLM 5.2
55.6
WeirdML
GLM 5.2 leads by +1.3
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
Claude Sonnet 5
68.8
GLM 5.2
70.1
Full benchmark table
| Benchmark | Claude Sonnet 5 | GLM 5.2 |
|---|---|---|
APEX-Agents APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments. | 54.5 | 45.2 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 42.1 | 17.9 |
Deepswe | 53.9 | 43.8 |
Dtbench | 87.5 | 89.3 |
Frontiercode | 42.7 | 24.5 |
FrontierMath-Tier-4-v2-Private | 29.3 | 29.3 |
FrontierMath-Tiers-1-3-v2-Private | 65.6 | 59.2 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 87.4 | 89.1 |
Lmca | 58.9 | 53.9 |
Mystery Game Puzzles | 28.4 | 10.8 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 94.7 | 86.4 |
Proofbench | 77.0 | 35.0 |
SimpleBench SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking. | 52.7 | 50.6 |
SimpleQA Verified SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information. | 33.7 | 34.2 |
Surface Evolver Bench | 60.0 | 55.6 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 68.8 | 70.1 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $2.00 | $10.00 | 1.0M tokens (~500 books) | $40.00 | |
| $0.02 | $7.00 | 1.0M tokens (~524 books) | $17.65 |