Compare · ModelsLive · 2 picked · head to head
GLM 5.3 vs gpt-oss-120b
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
GLM 5.3 wins on 16/16 benchmarks
GLM 5.3 wins 16 of 16 shared benchmarks. Leads in speed · agentic · knowledge.
Category leads
speed·GLM 5.3agentic·GLM 5.3knowledge·GLM 5.3general·GLM 5.3math·GLM 5.3coding·GLM 5.3
Hype vs Reality
Attention vs performance
GLM 5.3
#73 by perf·#3 by attention
gpt-oss-120b
#140 by perf·no signal
Best value
gpt-oss-120b
1.7x better value than GLM 5.3
GLM 5.3
77.1 pts/$
$0.73/M
gpt-oss-120b
131.5 pts/$
$0.38/M
Vendor risk
Who is behind the model
z-ai
private · undisclosed
OpenAI
$840.0B·Tier 1
Head to head
16 benchmarks · 2 models
GLM 5.3gpt-oss-120b
Artificial Analysis · CritPt
GLM 5.3 leads by +18.0
GLM 5.3
19.1
gpt-oss-120b
1.1
Artificial Analysis · GDPval
GLM 5.3 leads by +52.1
GLM 5.3
57.5
gpt-oss-120b
5.4
Artificial Analysis · GPQA Diamond
GLM 5.3 leads by +13.5
GLM 5.3
91.7
gpt-oss-120b
78.2
Artificial Analysis · Humanity's Last Exam
GLM 5.3 leads by +22.7
GLM 5.3
42.3
gpt-oss-120b
19.6
Artificial Analysis · Long Context Reasoning
GLM 5.3 leads by +27.7
GLM 5.3
79.7
gpt-oss-120b
52.0
Artificial Analysis · Quality Index
GLM 5.3 leads by +33.2
GLM 5.3
44.8
gpt-oss-120b
11.6
Artificial Analysis · SciCode
GLM 5.3 leads by +25.0
GLM 5.3
59.0
gpt-oss-120b
34.0
APEX-Agents
GLM 5.3 leads by +51.9
APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments.
GLM 5.3
56.6
gpt-oss-120b
4.7
Chess Puzzles
GLM 5.3 leads by +17.9
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
GLM 5.3
33.7
gpt-oss-120b
15.8
Dtbench
GLM 5.3 leads by +19.0
GLM 5.3
79.5
gpt-oss-120b
60.5
GPQA diamond
GLM 5.3 leads by +23.9
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
GLM 5.3
91.5
gpt-oss-120b
67.7
Lmca
GLM 5.3 leads by +39.3
GLM 5.3
65.3
gpt-oss-120b
26.1
Mystery Game Puzzles
GLM 5.3 leads by +26.2
GLM 5.3
26.2
gpt-oss-120b
0.0
OTIS Mock AIME 2024-2025
GLM 5.3 leads by +2.2
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
GLM 5.3
91.1
gpt-oss-120b
88.9
SimpleQA Verified
GLM 5.3 leads by +27.1
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
GLM 5.3
41.0
gpt-oss-120b
13.9
WeirdML
GLM 5.3 leads by +27.2
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
GLM 5.3
75.4
gpt-oss-120b
48.2
Full benchmark table
| Benchmark | GLM 5.3 | gpt-oss-120b |
|---|---|---|
Artificial Analysis · CritPt | 19.1 | 1.1 |
Artificial Analysis · GDPval | 57.5 | 5.4 |
Artificial Analysis · GPQA Diamond | 91.7 | 78.2 |
Artificial Analysis · Humanity's Last Exam | 42.3 | 19.6 |
Artificial Analysis · Long Context Reasoning | 79.7 | 52.0 |
Artificial Analysis · Quality Index | 44.8 | 11.6 |
Artificial Analysis · SciCode | 59.0 | 34.0 |
APEX-Agents APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments. | 56.6 | 4.7 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 33.7 | 15.8 |
Dtbench | 79.5 | 60.5 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 91.5 | 67.7 |
Lmca | 65.3 | 26.1 |
Mystery Game Puzzles | 26.2 | 0.0 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 91.1 | 88.9 |
SimpleQA Verified SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information. | 41.0 | 13.9 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 75.4 | 48.2 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.07 | $1.40 | 1.0M tokens (~524 books) | $4.02 | |
| $0.15 | $0.60 | 131K tokens (~66 books) | $2.62 |