DeepSeek V4 Pro 0813 vs DeepSeek V4 Pro

Change models
IndexAll categories, out of 100.
64.0 ±4.6 best
60.3 ±3.8
GapIndex points below the top.
Top
−3.7 · no clear gap
CoverageIndex weight that has results.
85%
85%
SpeedOutput tokens per second.
93/s
96/s best
EloLMArena rating and rank.
—
1451
ReleasedMonth of release.
2026-08
2026-08
Input / 1MUS dollars per 1M input tokens.
$1.32
$0.90 best
Output / 1MUS dollars per 1M output tokens.
$3.96
$1.50 best
Batch output / 1MBatch price per 1M output tokens.
$1.98
—
1K chats2K tokens in, 500 out, each.
$4.62
$2.55 best
1K documents20K tokens in, 1K out, each.
$30.36
$19.50 best
1K agent tasks200K tokens in, 20K out, each.
$343.20
$210.00 best
ContextMaximum tokens in one request.
1.05M
1.05M
AgenticMulti-step tasks with tools.
58.5
61.4 best
CodingCode writing and repair.
65.3 best
60.0
ReasoningLogic problems and puzzles.
67.5 best
58.6
MultimodalTasks with images and text.
—
—
KnowledgeFacts and expert knowledge.
62.0
62.4 best
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
62.7 best
56.0
MathMath problems.
64.3 best
56.7

Benchmarks

34 shared
Only DeepSeek V4 Pro 0813: 12Only DeepSeek V4 Pro: 49
DeepSeek V4 Pro 0813
DeepSeek V4 Pro
91.7% best34 / 220
89.6%48 / 220
98.6% best30 / 203
96.7%39 / 203
82.4% best72 / 149
80.3%87 / 149
TaxEval v2Finance
73.1% best52 / 143
72.1%68 / 143
87.5%13 / 142
87.5%13 / 142
92.4% best16 / 135
89.4%30 / 135
MMLU ProKnowledge
87.0%35 / 135
87.2% best32 / 135
CorpFin v2Finance
65.4% best27 / 131
61.4%52 / 131
82.3% best30 / 130
49.9%73 / 130
MedScribeHealthcare
80.2% best61 / 111
75.1%83 / 111
MedCodeHealthcare
42.5% best51 / 108
40.5%65 / 108
40.9% best38 / 98
23.1%73 / 98
52.9% best29 / 94
47.0%40 / 94
Chess PuzzlesReasoning
46.0% best29 / 93
19.0%71 / 93
64.6% best42 / 93
45.3%61 / 93
26.8% best43 / 89
25.2%47 / 89
SWE-benchCoding
96.4% best2 / 87
77.4%35 / 87
43.0% best20 / 83
17.0%54 / 83
79.2% best19 / 80
74.5%41 / 80
85.8% best35 / 80
82.7%51 / 80