Inference
EN

Claude Opus 4.7 vs DeepSeek V4 Flash 0731

Change models
Claude Opus 4.7Anthropic
DeepSeek V4 Flash 0731DeepSeek
IndexAll categories, out of 100.
62.0 ±3.0 best
61.9 ±3.8
GapIndex points below the top.
Top
−0.1 · no clear gap
CoverageIndex weight that has results.
95% best
85%
SpeedOutput tokens per second.
47/s
61/s best
EloLMArena rating and rank.
1483
—
ReleasedMonth of release.
2026-04
2026-07
Input / 1MUS dollars per 1M input tokens.
$5.00
$0.043 best
Output / 1MUS dollars per 1M output tokens.
$25.00
$0.30 best
Batch output / 1MBatch price per 1M output tokens.
$12.50
$0.144 best
1K chats2K tokens in, 500 out, each.
$22.50
$0.237 best
1K documents20K tokens in, 1K out, each.
$125.00
$1.17 best
1K agent tasks200K tokens in, 20K out, each.
$1,500.00
$14.68 best
ContextMaximum tokens in one request.
1M
1.05M best
AgenticMulti-step tasks with tools.
60.6 best
59.9
CodingCode writing and repair.
64.5 best
62.8
ReasoningLogic problems and puzzles.
61.1
65.5 best
MultimodalTasks with images and text.
61.3
—
KnowledgeFacts and expert knowledge.
61.8 best
58.6
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
45.7
59.2 best
MathMath problems.
64.3 best
61.5

Benchmarks

36 shared
Only Claude Opus 4.7: 29Only DeepSeek V4 Flash 0731: 49
Claude Opus 4.7
DeepSeek V4 Flash 0731
Artificial Analysis Humanity's Last ExamKnowledge
33.3%70 / 222
38.6% best53 / 222
Artificial Analysis Long Context ReasoningReasoning
75.7%92 / 217
79.7% best45 / 217
GPQA diamond · Epoch AIKnowledge
86.4%66 / 217
91.0% best33 / 217
Critical Physics TasksReasoning
5.1%83 / 214
16.6% best46 / 214
OTIS Mock AIME 2024-2025Math
86.7%69 / 200
94.4% best47 / 200
Artificial Analysis GPQA DiamondKnowledge
88.5%50 / 199
90.8% best30 / 199
Artificial Analysis Omniscience AccuracyKnowledge
44.7% best39 / 164
40.4%49 / 164
LegalBenchLegal
85.3% best22 / 149
77.7%107 / 149
TaxEval v2Finance
75.3% best19 / 143
70.7%88 / 143
LiveCodeBenchCoding
85.1%35 / 142
87.3% best16 / 142
GPQA Diamond · Vals AIKnowledge
90.2% best25 / 135
89.9%27 / 135
MMLU ProKnowledge
89.9% best8 / 135
86.2%45 / 135
CorpFin v2Finance
66.1% best21 / 131
61.8%51 / 131
Vibe Code Bench v1.1Coding
71.0%47 / 123
74.7% best43 / 123
MedScribeHealthcare
83.0% best48 / 109
80.4%57 / 109
MedCodeHealthcare
54.9% best7 / 106
41.4%53 / 106
SimpleQA VerifiedKnowledge
51.7% best28 / 91
33.6%63 / 91
Chess PuzzlesReasoning
41.0% best34 / 90
32.0%49 / 90
FrontierMath-Tiers-1-3-v2-PrivateMath
70.2% best29 / 90
57.5%46 / 90
Legal Research BenchLegal
38.5% best39 / 90
30.3%54 / 90