Claude Haiku 5.5 vs GLM-5.3-Flash

Change models
GLM-5.3-FlashZ.AIBudget
IndexAll categories, out of 100.
65.3 ±2.7 best
62.4 ±2.1
GapIndex points below the top.
Top
−2.9 · no clear gap
CoverageIndex weight that has results.
95%
100% best
SpeedOutput tokens per second.
167/s best
157/s
EloLMArena rating and rank.
1398
1471 best
ReleasedMonth of release.
2026-10
2026-08
Input / 1MUS dollars per 1M input tokens.
$0.10
$0.04 best
Output / 1MUS dollars per 1M output tokens.
$0.50
$0.078 best
Batch output / 1MBatch price per 1M output tokens.
$0.25
—
1K chats2K tokens in, 500 out, each.
$0.45
$0.119 best
1K documents20K tokens in, 1K out, each.
$2.50
$0.876 best
1K agent tasks200K tokens in, 20K out, each.
$30.00
$9.55 best
ContextMaximum tokens in one request.
1M
1.05M best
AgenticMulti-step tasks with tools.
67.4
68.7 best
CodingCode writing and repair.
69.6 best
62.2
ReasoningLogic problems and puzzles.
66.7 best
58.2
MultimodalTasks with images and text.
59.9
64.1 best
KnowledgeFacts and expert knowledge.
54.3
57.2 best
MultilingualTasks in many languages.
—
63.8
InstructionTasks with strict prompt rules.
37.7 best
34.2
MathMath problems.
67.2 best
56.6

Benchmarks

51 shared
Only Claude Haiku 5.5: 37Only GLM-5.3-Flash: 41
Claude Haiku 5.5
GLM-5.3-Flash
89.6%48 / 220
90.2% best43 / 220
98.9% best25 / 203
93.9%52 / 203
61.3%42 / 137
65.8% best36 / 137
89.0%42 / 132
91.0% best36 / 132
90.4% best5 / 130
30.8%89 / 130
64.4%34 / 127
66.8% best32 / 127
90.1%50 / 116
92.5% best44 / 116
MedScribeHealthcare
89.5% best7 / 111
88.9%10 / 111
43.8%28 / 98
45.2% best25 / 98
Chess PuzzlesReasoning
55.0% best19 / 93
21.0%62 / 93
75.1% best26 / 93
55.8%52 / 93
31.9% best30 / 89
31.6%31 / 89
30.0% best33 / 83
8.0%69 / 83
52.4%79 / 80
76.4% best35 / 80
83.5% best45 / 80
77.6%65 / 80
54.1%28 / 79
57.9% best13 / 79
44.6% best20 / 77
20.5%52 / 77
EMBFinance
59.0% best37 / 75
55.9%48 / 75
46.3% best27 / 75
17.1%55 / 75
36.4%26 / 68
37.4% best25 / 68