Muse Spark 1.3 vs Qwen4 Max

Change models
Qwen4 MaxAlibaba
IndexAll categories, out of 100.
69.6 ±4.0
72.0 ±8.0 best
GapIndex points below the top.
−2.4 · no clear gap
Top
CoverageIndex weight that has results.
85% best
65%
SpeedOutput tokens per second.
147/s best
40/s
EloLMArena rating and rank.
1492
—
ReleasedMonth of release.
2026-09
2026-11
Input / 1MUS dollars per 1M input tokens.
$1.25 best
$1.50
Output / 1MUS dollars per 1M output tokens.
$4.25 best
$4.50
1K chats2K tokens in, 500 out, each.
$4.63 best
$5.25
1K documents20K tokens in, 1K out, each.
$29.25 best
$34.50
1K agent tasks200K tokens in, 20K out, each.
$335.00 best
$390.00
ContextMaximum tokens in one request.
1.05M best
1M
AgenticMulti-step tasks with tools.
70.0
75.2 best
CodingCode writing and repair.
68.1 best
67.0
ReasoningLogic problems and puzzles.
67.8
—
MultimodalTasks with images and text.
—
72.1
KnowledgeFacts and expert knowledge.
67.2
70.7 best
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
83.8
—
MathMath problems.
68.2
—

Benchmarks

7 shared
Only Muse Spark 1.3: 52Only Qwen4 Max: 12
Muse Spark 1.3
Qwen4 Max
43.6%43 / 167
43.7% best42 / 167
82.9%25 / 125
86.7% best14 / 125
40.9%35 / 92
41.3% best31 / 92
37.9%16 / 84
40.4% best12 / 84
10.6%47 / 63
48.5% best14 / 63
DeepSWEAgentic
75.4%7 / 61
75.7% best5 / 61
1.026 / 56
8.4 best15 / 56