Inference
EN

Claude Fable 5.2 vs Muse Wonder

Change models
Claude Fable 5.2AnthropicSmartest
Muse WonderMeta
IndexAll categories, out of 100.
85.6 ±3.7 best
73.1 ±4.0
GapIndex points below the top.
Top
−12.5
CoverageIndex weight that has results.
90% best
85%
SpeedOutput tokens per second.
64/s
80/s best
EloLMArena rating and rank.
—
1503
ReleasedMonth of release.
2026-11
2026-10
Input / 1MUS dollars per 1M input tokens.
$8.00
$4.00 best
Output / 1MUS dollars per 1M output tokens.
$40.00
$20.00 best
Batch output / 1MBatch price per 1M output tokens.
$20.00
—
1K chats2K tokens in, 500 out, each.
$36.00
$18.00 best
1K documents20K tokens in, 1K out, each.
$200.00
$100.00 best
1K agent tasks200K tokens in, 20K out, each.
$2,400.00
$1,200.00 best
ContextMaximum tokens in one request.
1M
1.05M best
AgenticMulti-step tasks with tools.
82.2 best
73.0
CodingCode writing and repair.
86.2 best
72.3
ReasoningLogic problems and puzzles.
86.5 best
73.1
MultimodalTasks with images and text.
82.0
—
KnowledgeFacts and expert knowledge.
87.1 best
68.6
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
—
74.4
MathMath problems.
79.3 best
73.5

Benchmarks

33 shared
Only Claude Fable 5.2: 38Only Muse Wonder: 29
Claude Fable 5.2
Muse Wonder
Artificial Analysis Humanity's Last ExamKnowledge
74.3% best1 / 222
50.1%12 / 222
Artificial Analysis Long Context ReasoningReasoning
83.3% best13 / 217
80.7%37 / 217
Critical Physics TasksReasoning
35.4% best1 / 214
28.3%15 / 214
OTIS Mock AIME 2024-2025Math
99.7% best14 / 200
99.4%17 / 200
Artificial Analysis Omniscience AccuracyKnowledge
79.1% best1 / 164
38.0%59 / 164
Vibe Code Bench v1.1Coding
93.8% best1 / 123
83.1%22 / 123
Artificial Analysis SciCodeCoding
67.7% best1 / 114
60.3%10 / 114
FrontierMath-Tiers-1-3-v2-PrivateMath
95.1% best1 / 90
82.1%18 / 90
Legal Research BenchLegal
46.2%19 / 90
50.5% best7 / 90
Tax Agent BenchFinance
32.1%23 / 82
36.3% best15 / 82
Mystery Game PuzzlesReasoning
90.0% best1 / 80
51.0%15 / 80
FrontierMath-Tier-4-v2-PrivateMath
100.0% best1 / 72
70.7%17 / 72
Terminal-Bench 4.0Agentic
65.2% best1 / 61
49.5%11 / 61
DeepSWEAgentic
74.3%9 / 59
78.8% best2 / 59
Agent Arena steerabilityAgentic
33.0 best1 / 54
2.622 / 54
Agent Arena task outcomeAgentic
22.0 best1 / 54
5.413 / 54
ProofBench v1.1Math
100.0% best1 / 54
84.0%15 / 54
Terminal-Bench-ScienceScience
55.7% best2 / 44
22.9%11 / 44
Furniture AssemblyReasoning
90.0% best1 / 37
54.2%13 / 37
HLAB-Vals-1.1Legal
5.0%27 / 35
23.8% best4 / 35