DeepSeek V4 Pro vs Gemini 3.1 Pro Preview

Change models
IndexAll categories, out of 100.
60.3 ±3.8
60.7 ±3.6 best
GapIndex points below the top.
−0.4 · no clear gap
Top
CoverageIndex weight that has results.
85%
95% best
SpeedOutput tokens per second.
96/s best
74/s
EloLMArena rating and rank.
1451
1480 best
ReleasedMonth of release.
2026-08
2026-02
Input / 1MUS dollars per 1M input tokens.
$0.90 best
$2.00
Output / 1MUS dollars per 1M output tokens.
$1.50 best
$12.00
Batch output / 1MBatch price per 1M output tokens.
—
$6.00
1K chats2K tokens in, 500 out, each.
$2.55 best
$10.00
1K documents20K tokens in, 1K out, each.
$19.50 best
$52.00
1K agent tasks200K tokens in, 20K out, each.
$210.00 best
$640.00
ContextMaximum tokens in one request.
1.05M
1.05M
AgenticMulti-step tasks with tools.
61.4 best
52.1
CodingCode writing and repair.
60.0 best
58.2
ReasoningLogic problems and puzzles.
58.6
68.6 best
MultimodalTasks with images and text.
—
69.0
KnowledgeFacts and expert knowledge.
62.4
64.3 best
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
56.0
84.4 best
MathMath problems.
56.7
59.6 best

Benchmarks

31 shared
Only DeepSeek V4 Pro: 52Only Gemini 3.1 Pro Preview: 25
DeepSeek V4 Pro
Gemini 3.1 Pro Preview
80.3%87 / 149
87.4% best4 / 149
TaxEval v2Finance
72.1%68 / 143
72.9% best56 / 143
87.5%13 / 142
88.5% best6 / 142
89.4%30 / 135
95.5% best1 / 135
MMLU ProKnowledge
87.2%32 / 135
91.0% best4 / 135
CorpFin v2Finance
61.4%52 / 131
64.5% best35 / 131
49.9% best73 / 130
32.0%87 / 130
MedScribeHealthcare
75.1%83 / 111
76.1% best79 / 111
MedCodeHealthcare
40.5%65 / 108
59.1% best2 / 108
23.1% best73 / 98
20.7%74 / 98
25.2% best47 / 89
9.3%75 / 89
SWE-benchCoding
77.4%35 / 87
78.8% best28 / 87
74.5%41 / 80
78.5% best23 / 80
82.7%51 / 80
84.0% best42 / 80
44.1% best58 / 79
43.0%59 / 79
26.2% best46 / 77
17.3%54 / 77
EMBFinance
51.6%55 / 75
52.6% best54 / 75
50.2%58 / 75
70.8% best27 / 75
11.1% best50 / 68
2.5%61 / 68
42.6%57 / 66
44.1% best54 / 66