GLM-4.5 vs o4-mini

Change models
o4-miniOpenAI
IndexAll categories, out of 100.
45.3 ±9.2 best
44.5 ±3.0
GapIndex points below the top.
Top
−0.8 · no clear gap
CoverageIndex weight that has results.
65%
100% best
SpeedOutput tokens per second.
40/s
122/s best
EloLMArena rating and rank.
1430 best
1353
ReleasedMonth of release.
2025-07
2025-04
Input / 1MUS dollars per 1M input tokens.
$0.60 best
$1.10
Output / 1MUS dollars per 1M output tokens.
$2.20 best
$4.40
1K chats2K tokens in, 500 out, each.
$2.30 best
$4.40
1K documents20K tokens in, 1K out, each.
$14.20 best
$26.40
1K agent tasks200K tokens in, 20K out, each.
$164.00 best
$308.00
ContextMaximum tokens in one request.
131K
200K best
AgenticMulti-step tasks with tools.
47.6 best
39.5
CodingCode writing and repair.
43.0
44.9 best
ReasoningLogic problems and puzzles.
—
44.6
MultimodalTasks with images and text.
—
51.5
KnowledgeFacts and expert knowledge.
44.8 best
40.9
MultilingualTasks in many languages.
—
35.4
InstructionTasks with strict prompt rules.
—
52.5
MathMath problems.
47.8
49.8 best

Benchmarks

15 shared
Only GLM-4.5: 1Only o4-mini: 42
GLM-4.5
o4-mini
75.6%114 / 149
79.2% best97 / 149
TaxEval v2Finance
72.4%62 / 143
74.8% best29 / 143
67.4%96 / 142
82.2% best53 / 142
72.2%88 / 135
74.5% best83 / 135
MMLU ProKnowledge
81.2% best82 / 135
80.6%86 / 135
CorpFin v2Finance
61.0% best60 / 131
59.0%79 / 131
AIMEMath
86.7% best31 / 95
83.7%41 / 95
MedQAHealthcare
90.0%50 / 95
96.0% best8 / 95
MGSMMultilingual
90.8%37 / 74
93.4% best9 / 74
59.2%32 / 65
64.6% best23 / 65
0.0%33 / 64
30.0% best22 / 64
IOI v1Coding
2.9%53 / 61
4.8% best45 / 61
94.0%14 / 59
94.2% best10 / 59
54.2% best18 / 46
45.0%24 / 46
0.0%3 / 36
0.0%3 / 36