Inference
EN

DeepSeek V4 Flash 0731 vs Mistral Large 4

Change models
DeepSeek V4 Flash 0731DeepSeek
Mistral Large 4Mistral
IndexAll categories, out of 100.
61.9 ±3.8 best
60.7 ±4.1
GapIndex points below the top.
Top
−1.2 · no clear gap
CoverageIndex weight that has results.
85%
85%
SpeedOutput tokens per second.
61/s
113/s best
EloLMArena rating and rank.
—
1414
ReleasedMonth of release.
2026-07
2026-10
Input / 1MUS dollars per 1M input tokens.
$0.043 best
$0.70
Output / 1MUS dollars per 1M output tokens.
$0.30 best
$2.10
Batch output / 1MBatch price per 1M output tokens.
$0.144
—
1K chats2K tokens in, 500 out, each.
$0.237 best
$2.45
1K documents20K tokens in, 1K out, each.
$1.17 best
$16.10
1K agent tasks200K tokens in, 20K out, each.
$14.68 best
$182.00
ContextMaximum tokens in one request.
1.05M best
1M
AgenticMulti-step tasks with tools.
59.9
66.3 best
CodingCode writing and repair.
62.8 best
62.7
ReasoningLogic problems and puzzles.
65.5 best
59.2
MultimodalTasks with images and text.
—
—
KnowledgeFacts and expert knowledge.
58.6 best
43.7
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
59.2 best
57.7
MathMath problems.
61.5
61.6 best

Benchmarks

34 shared
Only DeepSeek V4 Flash 0731: 51Only Mistral Large 4: 26
DeepSeek V4 Flash 0731
Mistral Large 4
GPQA diamond · Epoch AIKnowledge
91.0% best33 / 217
89.1%50 / 217
Critical Physics TasksReasoning
16.6% best46 / 214
10.0%65 / 214
OTIS Mock AIME 2024-2025Math
94.4%47 / 200
98.9% best22 / 200
Artificial Analysis Omniscience AccuracyKnowledge
40.4% best49 / 164
26.0%99 / 164
GDPval-AA normalizedAgentic
46.7% best35 / 133
46.1%37 / 133
Vibe Code Bench v1.1Coding
74.7%43 / 123
78.4% best35 / 123
MedScribeHealthcare
80.4%57 / 109
80.4%57 / 109
MedCodeHealthcare
41.4% best53 / 106
40.7%61 / 106
SimpleQA VerifiedKnowledge
33.6% best63 / 91
21.2%77 / 91
Chess PuzzlesReasoning
32.0% best49 / 90
31.0%50 / 90
FrontierMath-Tiers-1-3-v2-PrivateMath
57.5%46 / 90
58.6% best45 / 90
Legal Research BenchLegal
30.3%54 / 90
31.7% best51 / 90
Tax Agent BenchFinance
28.1%38 / 82
30.4% best33 / 82
LiveBench Data AnalysisReasoning
79.3% best15 / 80
76.5%33 / 80
LiveBench ReasoningReasoning
86.6% best32 / 80
83.9%43 / 80
Mystery Game PuzzlesReasoning
34.0% best25 / 80
18.5%48 / 80
Finance Agent (v2)Finance
49.5%47 / 79
54.7% best26 / 79
Code MigrationCoding
38.6% best32 / 77
30.6%41 / 77
EMBFinance
57.0% best42 / 75
56.2%47 / 75
FrontierMath-Tier-4-v2-PrivateMath
24.4%46 / 72
26.8% best39 / 72