Muse Spark Open vs Qwen3.8 Max

Change models
IndexAll categories, out of 100.
64.1 ±3.9
64.4 ±2.5 best
GapIndex points below the top.
−0.3 · no clear gap
Top
CoverageIndex weight that has results.
90%
95% best
SpeedOutput tokens per second.
141/s best
41/s
EloLMArena rating and rank.
—
1485
ReleasedMonth of release.
2026-10
2026-08
Input / 1MUS dollars per 1M input tokens.
$1.25
—
Output / 1MUS dollars per 1M output tokens.
$4.25
—
1K chats2K tokens in, 500 out, each.
$4.63
—
1K documents20K tokens in, 1K out, each.
$29.25
—
1K agent tasks200K tokens in, 20K out, each.
$335.00
—
ContextMaximum tokens in one request.
1.05M best
1M
AgenticMulti-step tasks with tools.
63.7 best
62.6
CodingCode writing and repair.
62.2
63.3 best
ReasoningLogic problems and puzzles.
61.5
65.2 best
MultimodalTasks with images and text.
69.1 best
65.9
KnowledgeFacts and expert knowledge.
64.6 best
61.8
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
—
70.0
MathMath problems.
63.7
66.4 best

Benchmarks

28 shared
Only Muse Spark Open: 17Only Qwen3.8 Max: 86
Muse Spark Open
Qwen3.8 Max
88.6%56 / 220
92.7% best28 / 220
98.1%34 / 203
99.4% best20 / 203
76.2% best47 / 130
64.7%62 / 130
38.5%44 / 98
47.6% best18 / 98
56.6% best25 / 94
45.8%44 / 94
69.8%33 / 93
74.7% best27 / 93
26.3%45 / 89
32.1% best26 / 89
60.2%72 / 80
78.4% best24 / 80
73.7%71 / 80
88.2% best24 / 80
61.3% best3 / 79
50.6%43 / 79
21.2%51 / 77
24.0% best49 / 77
EMBFinance
53.7%51 / 75
60.1% best35 / 75
31.7%35 / 75
46.3% best27 / 75
6.6%55 / 68
34.3% best27 / 68
DeepSWEAgentic
60.0% best44 / 64
56.6%50 / 64
0.0%29 / 55
0.0%29 / 55
73.6% best37 / 52
60.3%46 / 52
60.3%47 / 52
78.6% best17 / 52
80.3% best17 / 52
69.9%33 / 52
18.9%42 / 51
42.3% best19 / 51