Inference
EN

Gemini 3.8 Flash vs Muse Wonder Full

Change models
Gemini 3.8 FlashGoogle
Muse Wonder FullMeta
IndexAll categories, out of 100.
67.8 ±2.5
77.3 ±5.3 best
GapIndex points below the top.
−9.5
Top
CoverageIndex weight that has results.
95% best
80%
SpeedOutput tokens per second.
68/s
—
EloLMArena rating and rank.
1500
—
ReleasedMonth of release.
2026-09
2026-10
Input / 1MUS dollars per 1M input tokens.
$0.825
—
Output / 1MUS dollars per 1M output tokens.
$4.13
—
Batch output / 1MBatch price per 1M output tokens.
$1.88
—
1K chats2K tokens in, 500 out, each.
$3.71
—
1K documents20K tokens in, 1K out, each.
$20.63
—
1K agent tasks200K tokens in, 20K out, each.
$247.50
—
ContextMaximum tokens in one request.
1.05M best
1M
AgenticMulti-step tasks with tools.
65.3
78.2 best
CodingCode writing and repair.
66.2
77.3 best
ReasoningLogic problems and puzzles.
68.3
75.6 best
MultimodalTasks with images and text.
67.1
—
KnowledgeFacts and expert knowledge.
70.9
71.7 best
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
91.3
—
MathMath problems.
64.3
79.6 best

Benchmarks

36 shared
Only Gemini 3.8 Flash: 70Only Muse Wonder Full: 6
Gemini 3.8 Flash
Muse Wonder Full
Artificial Analysis Humanity's Last ExamKnowledge
47.8%20 / 222
53.6% best9 / 222
Artificial Analysis Long Context ReasoningReasoning
81.3%34 / 217
83.7% best10 / 217
Critical Physics TasksReasoning
18.3%34 / 214
32.9% best2 / 214
Artificial Analysis Omniscience AccuracyKnowledge
50.5% best19 / 164
41.0%46 / 164
LegalBenchLegal
87.0%7 / 149
87.4% best4 / 149
GDPval-AA normalizedAgentic
46.8%34 / 133
67.9% best2 / 133
Vibe Code Bench v1.1Coding
78.7%34 / 123
86.0% best14 / 123
Artificial Analysis SciCodeCoding
56.6%23 / 114
62.6% best4 / 114
MedScribeHealthcare
84.5%34 / 109
89.3% best8 / 109
MedCodeHealthcare
48.1%31 / 106
49.9% best18 / 106
Legal Research BenchLegal
38.9%38 / 90
53.4% best6 / 90
Tax Agent BenchFinance
32.1%23 / 82
43.0% best8 / 82
Finance Agent (v2)Finance
61.4% best2 / 79
59.1%6 / 79
Code MigrationCoding
36.5%35 / 77
61.0% best7 / 77
EMBFinance
72.2% best8 / 75
71.9%10 / 75
ProgramBenchCoding
1.0% best16 / 65
0.5%19 / 65
Terminal-Bench 4.0Agentic
19.7%34 / 61
58.1% best6 / 61
DeepSWEAgentic
73.8%12 / 59
79.0% best1 / 59
Agent Arena steerabilityAgentic
−1.333 / 54
9.6 best12 / 54
Agent Arena task outcomeAgentic
1.329 / 54
11.1 best9 / 54