Claude Opus 4.5 vs Qwen3.5-122B-A10B

Change models
IndexAll categories, out of 100.
51.4 ±2.1 best
51.2 ±4.3
GapIndex points below the top.
Top
−0.2 · no clear gap
CoverageIndex weight that has results.
100% best
90%
SpeedOutput tokens per second.
37/s
129/s best
EloLMArena rating and rank.
1451 best
1418
ReleasedMonth of release.
2025-11
2026-03
Input / 1MUS dollars per 1M input tokens.
$5.00
$0.26 best
Output / 1MUS dollars per 1M output tokens.
$25.00
$2.08 best
Batch output / 1MBatch price per 1M output tokens.
$12.50
—
1K chats2K tokens in, 500 out, each.
$22.50
$1.56 best
1K documents20K tokens in, 1K out, each.
$125.00
$7.28 best
1K agent tasks200K tokens in, 20K out, each.
$1,500.00
$93.60 best
ContextMaximum tokens in one request.
200K
262K best
AgenticMulti-step tasks with tools.
54.0 best
50.8
CodingCode writing and repair.
54.3 best
49.8
ReasoningLogic problems and puzzles.
51.2 best
49.7
MultimodalTasks with images and text.
41.7
53.1 best
KnowledgeFacts and expert knowledge.
54.0 best
52.4
MultilingualTasks in many languages.
55.8 best
51.2
InstructionTasks with strict prompt rules.
41.0
54.7 best
MathMath problems.
50.6
—

Benchmarks

23 shared
Only Claude Opus 4.5: 79Only Qwen3.5-122B-A10B: 5
Claude Opus 4.5
Qwen3.5-122B-A10B
13.2%127 / 216
25.2% best95 / 216
70.7%105 / 211
76.3% best84 / 211
0.3%132 / 207
0.6% best127 / 207
81.0%92 / 193
85.7% best69 / 193
87.0% best45 / 144
86.6%47 / 144
86.3%48 / 143
93.6% best28 / 143
43.0%89 / 140
75.7% best18 / 140
80.9% best13 / 114
72.0%62 / 114
71.2%69 / 109
75.0% best51 / 109
22.0% best44 / 83
17.0%54 / 83
89.5% best2 / 73
86.7%9 / 73
90.9%24 / 69
93.4% best14 / 69
70.6%45 / 58
76.9% best29 / 58
84.8% best21 / 57
78.9%31 / 57
68.5%48 / 50
77.2% best45 / 50
59.3% best18 / 40
49.4%30 / 40
66.3% best25 / 36
58.0%32 / 36
80.7%11 / 32
83.9% best6 / 32
LongBench v2Reasoning
64.4% best7 / 25
60.2%16 / 25
MathVisionMultimodal
74.3%24 / 25
86.2% best16 / 25