DeepSeek V4 Pro vs GLM-5.2

Change models
IndexAll categories, out of 100.
60.3 ±3.8 best
59.9 ±3.7
GapIndex points below the top.
Top
−0.4 · no clear gap
CoverageIndex weight that has results.
85%
85%
SpeedOutput tokens per second.
96/s
230/s best
EloLMArena rating and rank.
1451
1469 best
ReleasedMonth of release.
2026-08
2026-06
Input / 1MUS dollars per 1M input tokens.
$0.90
$0.62 best
Output / 1MUS dollars per 1M output tokens.
$1.50 best
$2.42
Batch output / 1MBatch price per 1M output tokens.
—
$0.989
1K chats2K tokens in, 500 out, each.
$2.55
$2.45 best
1K documents20K tokens in, 1K out, each.
$19.50
$14.82 best
1K agent tasks200K tokens in, 20K out, each.
$210.00
$172.40 best
ContextMaximum tokens in one request.
1.05M
1.05M
AgenticMulti-step tasks with tools.
61.4 best
60.7
CodingCode writing and repair.
60.0
63.3 best
ReasoningLogic problems and puzzles.
58.6 best
57.8
MultimodalTasks with images and text.
—
—
KnowledgeFacts and expert knowledge.
62.4 best
55.3
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
56.0 best
54.5
MathMath problems.
56.7
58.6 best

Benchmarks

54 shared
Only DeepSeek V4 Pro: 29Only GLM-5.2: 25
DeepSeek V4 Pro
GLM-5.2
89.6%48 / 220
91.9% best33 / 220
41.0%43 / 216
41.1% best42 / 216
80.3% best40 / 211
78.3%63 / 211
18.0%34 / 207
20.9% best25 / 207
96.7% best39 / 203
86.4%76 / 203
92.8% best14 / 193
89.5%43 / 193
49.1% best26 / 170
24.3%109 / 170
80.3%87 / 149
84.1% best39 / 149
90.1%31 / 144
91.2% best25 / 144
TaxEval v2Finance
72.1%68 / 143
73.3% best50 / 143
96.2%12 / 143
99.1% best1 / 143
87.5% best13 / 142
69.5%92 / 142
76.5% best11 / 140
73.3%28 / 140
89.4% best30 / 135
85.6%45 / 135
MMLU ProKnowledge
87.2% best32 / 135
86.7%39 / 135
54.5% best23 / 132
43.6%42 / 132
CorpFin v2Finance
61.4%52 / 131
66.1% best21 / 131
49.9%73 / 130
64.0% best63 / 130
51.0%48 / 113
51.2% best47 / 113
MedScribeHealthcare
75.1%83 / 111
83.5% best48 / 111