Inference
EN

MiMo-V2.6-Pro vs Muse Spark 1.3

Change models
MiMo-V2.6-ProXiaomi
Muse Spark 1.3Meta
IndexAll categories, out of 100.
70.2 ±4.7 best
70.1 ±4.0
GapIndex points below the top.
Top
−0.1 · no clear gap
CoverageIndex weight that has results.
80%
85% best
SpeedOutput tokens per second.
31/s
76/s best
EloLMArena rating and rank.
1492 best
1491
ReleasedMonth of release.
2026-09
2026-09
Input / 1MUS dollars per 1M input tokens.
$0.435 best
$1.25
Output / 1MUS dollars per 1M output tokens.
$0.87 best
$4.25
1K chats2K tokens in, 500 out, each.
$1.31 best
$4.63
1K documents20K tokens in, 1K out, each.
$9.57 best
$29.25
1K agent tasks200K tokens in, 20K out, each.
$104.40 best
$335.00
ContextMaximum tokens in one request.
1.05M
1.05M
AgenticMulti-step tasks with tools.
72.1 best
70.4
CodingCode writing and repair.
66.6
68.8 best
ReasoningLogic problems and puzzles.
68.3 best
68.1
MultimodalTasks with images and text.
—
—
KnowledgeFacts and expert knowledge.
64.2
67.5 best
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
—
84.4
MathMath problems.
71.3 best
68.6

Benchmarks

34 shared
Only MiMo-V2.6-Pro: 24Only Muse Spark 1.3: 25
MiMo-V2.6-Pro
Muse Spark 1.3
Artificial Analysis Humanity's Last ExamKnowledge
49.4% best14 / 222
48.7%15 / 222
Artificial Analysis Long Context ReasoningReasoning
86.3% best3 / 217
83.0%18 / 217
Critical Physics TasksReasoning
26.6% best19 / 214
24.9%21 / 214
Artificial Analysis Omniscience AccuracyKnowledge
34.8%67 / 164
43.6% best41 / 164
GDPval-AA normalizedAgentic
59.2% best9 / 133
59.0%10 / 133
Vibe Code Bench v1.1Coding
85.2% best16 / 123
82.9%23 / 123
Artificial Analysis SciCodeCoding
60.9% best8 / 114
58.8%14 / 114
Legal Research BenchLegal
47.1% best16 / 90
40.9%33 / 90
Terminal-Bench 2.1 · BenchLMAgentic
89.9% best4 / 86
88.8%7 / 86
Tax Agent BenchFinance
32.4%22 / 82
37.9% best14 / 82
LiveBench Data AnalysisReasoning
72.3%49 / 80
79.6% best13 / 80
LiveBench ReasoningReasoning
79.6%59 / 80
89.7% best15 / 80
Finance Agent (v2)Finance
57.3%16 / 79
58.9% best8 / 79
Code MigrationCoding
43.0% best26 / 77
27.6%44 / 77
EMBFinance
62.9% best30 / 75
62.7%31 / 75
Terminal-Bench 2.1 · Vals AIAgentic
67.8%32 / 75
72.3% best24 / 75
ProgramBenchCoding
0.5%19 / 65
0.5%19 / 65
Terminal-Bench 4.0Agentic
31.3% best24 / 61
10.6%45 / 61
DeepSWEAgentic
71.9%18 / 59
75.4% best6 / 59
Agent Arena steerabilityAgentic
1.525 / 54
2.6 best22 / 54