DeepSeek V4 Pro 0813 vs Muse Spark 1.2

Change models
IndexAll categories, out of 100.
64.0 ±4.6 best
63.9 ±3.3
GapIndex points below the top.
Top
−0.1 · no clear gap
CoverageIndex weight that has results.
85%
95% best
SpeedOutput tokens per second.
93/s
115/s best
EloLMArena rating and rank.
—
1484
ReleasedMonth of release.
2026-08
2026-08
Input / 1MUS dollars per 1M input tokens.
$1.32
$1.25 best
Output / 1MUS dollars per 1M output tokens.
$3.96 best
$4.25
Batch output / 1MBatch price per 1M output tokens.
$1.98
—
1K chats2K tokens in, 500 out, each.
$4.62 best
$4.63
1K documents20K tokens in, 1K out, each.
$30.36
$29.25 best
1K agent tasks200K tokens in, 20K out, each.
$343.20
$335.00 best
ContextMaximum tokens in one request.
1.05M
1.05M
AgenticMulti-step tasks with tools.
58.5
61.6 best
CodingCode writing and repair.
65.3 best
63.5
ReasoningLogic problems and puzzles.
67.5 best
63.0
MultimodalTasks with images and text.
—
66.4
KnowledgeFacts and expert knowledge.
62.0
64.4 best
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
62.7
75.3 best
MathMath problems.
64.3 best
63.6

Benchmarks

30 shared
Only DeepSeek V4 Pro 0813: 16Only Muse Spark 1.2: 27
DeepSeek V4 Pro 0813
Muse Spark 1.2
82.4%72 / 149
85.3% best22 / 149
TaxEval v2Finance
73.1%52 / 143
80.4% best1 / 143
MMLU ProKnowledge
87.0%35 / 135
88.3% best20 / 135
CorpFin v2Finance
65.4%27 / 131
70.9% best5 / 131
82.3% best30 / 130
79.1%37 / 130
MedScribeHealthcare
80.2%61 / 111
90.1% best6 / 111
MedCodeHealthcare
42.5%51 / 108
49.3% best24 / 108
40.9%38 / 98
43.8% best28 / 98
52.9%29 / 94
60.3% best21 / 94
Chess PuzzlesReasoning
46.0% best29 / 93
27.0%55 / 93
26.8%43 / 89
32.8% best23 / 89
SWE-benchCoding
96.4% best2 / 87
86.6%13 / 87
79.2% best19 / 80
76.5%33 / 80
85.8%35 / 80
90.0% best14 / 80
50.4%44 / 79
60.6% best4 / 79
41.5% best28 / 77
30.0%42 / 77
EMBFinance
52.8%53 / 75
57.0% best42 / 75
54.7%51 / 75
69.7% best28 / 75
14.1% best45 / 68
6.1%57 / 68
54.9%29 / 66
57.6% best20 / 66