A.X K2 vs Qwen3.5-122B-A10B

Change models
A.X K2SK Telecom
IndexAll categories, out of 100.
51.8 ±6.6
51.8 ±4.3
GapIndex points below the top.
Top
Top
CoverageIndex weight that has results.
85%
90% best
SpeedOutput tokens per second.
—
70/s
EloLMArena rating and rank.
—
1417
ReleasedMonth of release.
2026-07
2026-03
Input / 1MUS dollars per 1M input tokens.
—
$0.26
Output / 1MUS dollars per 1M output tokens.
—
$2.08
1K chats2K tokens in, 500 out, each.
—
$1.56
1K documents20K tokens in, 1K out, each.
—
$7.28
1K agent tasks200K tokens in, 20K out, each.
—
$93.60
ContextMaximum tokens in one request.
262K
262K
AgenticMulti-step tasks with tools.
46.3
51.3 best
CodingCode writing and repair.
51.2 best
50.3
ReasoningLogic problems and puzzles.
53.3 best
50.2
MultimodalTasks with images and text.
—
53.8
KnowledgeFacts and expert knowledge.
54.3 best
52.8
MultilingualTasks in many languages.
—
51.5
InstructionTasks with strict prompt rules.
59.9 best
55.4
MathMath problems.
55.1
—

Benchmarks

7 shared
Only A.X K2: 8Only Qwen3.5-122B-A10B: 21
A.X K2
Qwen3.5-122B-A10B
29.6% best82 / 222
25.2%99 / 222
66.0%133 / 217
76.3% best86 / 217
8.9% best72 / 214
0.6%132 / 214
85.7%72 / 199
85.7%72 / 199
23.0% best78 / 133
15.7%86 / 133
41.0% best83 / 114
39.7%89 / 114
84.0% best23 / 59
78.9%32 / 59