Inference
EN

DeepSeek V4 Flash 0731 vs Gemini 3.1 Pro Preview

Change models
DeepSeek V4 Flash 0731DeepSeek
Gemini 3.1 Pro PreviewGoogle
IndexAll categories, out of 100.
61.9 ±3.8 best
61.5 ±3.6
GapIndex points below the top.
Top
−0.4 · no clear gap
CoverageIndex weight that has results.
85%
95% best
SpeedOutput tokens per second.
61/s
77/s best
EloLMArena rating and rank.
—
1480
ReleasedMonth of release.
2026-07
2026-02
Input / 1MUS dollars per 1M input tokens.
$0.043 best
$2.00
Output / 1MUS dollars per 1M output tokens.
$0.30 best
$12.00
Batch output / 1MBatch price per 1M output tokens.
$0.144 best
$6.00
1K chats2K tokens in, 500 out, each.
$0.237 best
$10.00
1K documents20K tokens in, 1K out, each.
$1.17 best
$52.00
1K agent tasks200K tokens in, 20K out, each.
$14.68 best
$640.00
ContextMaximum tokens in one request.
1.05M
1.05M
AgenticMulti-step tasks with tools.
59.9 best
52.5
CodingCode writing and repair.
62.8 best
58.8
ReasoningLogic problems and puzzles.
65.5
69.8 best
MultimodalTasks with images and text.
—
69.9
KnowledgeFacts and expert knowledge.
58.6
65.0 best
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
59.2
86.6 best
MathMath problems.
61.5 best
60.4

Benchmarks

37 shared
Only DeepSeek V4 Flash 0731: 48Only Gemini 3.1 Pro Preview: 19
DeepSeek V4 Flash 0731
Gemini 3.1 Pro Preview
LegalBenchLegal
77.7%107 / 149
87.4% best4 / 149
TaxEval v2Finance
70.7%88 / 143
72.9% best56 / 143
LiveCodeBenchCoding
87.3%16 / 142
88.5% best6 / 142
GPQA Diamond · Vals AIKnowledge
89.9%27 / 135
95.5% best1 / 135
MMLU ProKnowledge
86.2%45 / 135
91.0% best4 / 135
ARC-AGI-2 (semi-private)Reasoning
61.4%36 / 131
77.1% best21 / 131
CorpFin v2Finance
61.8%51 / 131
64.5% best35 / 131
ARC-AGI-1 (semi-private)Reasoning
89.0%39 / 126
98.0% best4 / 126
Vibe Code Bench v1.1Coding
74.7% best43 / 123
32.0%80 / 123
ARC-AGI-2 (public eval)Reasoning
63.9%33 / 122
88.1% best18 / 122
ARC-AGI-1 (public eval)Reasoning
94.8%35 / 116
97.2% best21 / 116
MedScribeHealthcare
80.4% best57 / 109
76.1%77 / 109
MedCodeHealthcare
41.4%53 / 106
59.1% best2 / 106
Legal Research BenchLegal
30.3% best54 / 90
20.7%67 / 90
SWE-benchCoding
88.8% best11 / 87
78.8%28 / 87
Tax Agent BenchFinance
28.1% best38 / 82
9.3%68 / 82
LiveBench Data AnalysisReasoning
79.3% best15 / 80
78.5%23 / 80
LiveBench ReasoningReasoning
86.6% best32 / 80
84.0%42 / 80
Finance Agent (v2)Finance
49.5% best47 / 79
43.0%59 / 79
Code MigrationCoding
38.6% best32 / 77
17.3%54 / 77