GLM-5.3-Flash vs Grok 4.5

Change models
GLM-5.3-FlashZ.AIBudget
IndexAll categories, out of 100.
62.4 ±2.1
62.7 ±2.9 best
GapIndex points below the top.
−0.3 · no clear gap
Top
CoverageIndex weight that has results.
100% best
95%
SpeedOutput tokens per second.
157/s best
47/s
EloLMArena rating and rank.
1471 best
1448
ReleasedMonth of release.
2026-08
2026-07
Input / 1MUS dollars per 1M input tokens.
$0.04 best
$2.00
Output / 1MUS dollars per 1M output tokens.
$0.078 best
$6.00
1K chats2K tokens in, 500 out, each.
$0.119 best
$7.00
1K documents20K tokens in, 1K out, each.
$0.876 best
$46.00
1K agent tasks200K tokens in, 20K out, each.
$9.55 best
$520.00
ContextMaximum tokens in one request.
1.05M best
500K
AgenticMulti-step tasks with tools.
68.7 best
60.6
CodingCode writing and repair.
62.2
64.4 best
ReasoningLogic problems and puzzles.
58.2
62.4 best
MultimodalTasks with images and text.
64.1 best
47.3
KnowledgeFacts and expert knowledge.
57.2
64.9 best
MultilingualTasks in many languages.
63.8
—
InstructionTasks with strict prompt rules.
34.2
70.0 best
MathMath problems.
56.6
60.9 best

Benchmarks

51 shared
Only GLM-5.3-Flash: 41Only Grok 4.5: 23
GLM-5.3-Flash
Grok 4.5
90.2%43 / 220
93.4% best21 / 220
93.9%52 / 203
97.8% best36 / 203
83.9%45 / 149
86.0% best17 / 149
TaxEval v2Finance
75.6% best14 / 143
71.7%74 / 143
80.5%67 / 142
87.4% best15 / 142
65.8% best36 / 137
52.6%54 / 137
86.4%43 / 135
92.9% best12 / 135
MMLU ProKnowledge
86.1%49 / 135
89.2% best14 / 135
91.0% best36 / 132
85.7%54 / 132
30.8%89 / 130
69.0% best58 / 130
66.8% best32 / 127
58.2%46 / 127
92.5%44 / 116
94.8% best38 / 116
MedScribeHealthcare
88.9% best10 / 111
86.9%20 / 111
45.2% best25 / 98
38.0%49 / 98
Chess PuzzlesReasoning
21.0%62 / 93
35.0% best49 / 93
55.8%52 / 93
57.2% best50 / 93
MMMU ProMultimodal
86.0% best23 / 92
61.8%82 / 92
31.6% best31 / 89
24.9%48 / 89
84.3% best23 / 89
83.3%25 / 89
SWE-benchCoding
92.0% best10 / 87
86.6%13 / 87