Kimi K2.6 vs Beam 1.0 501B-A23B NVFP4

Change models
Kimi K2.6Moonshot AI
IndexAll categories, out of 100.
58.2 ±2.9
58.5 ±5.4 best
GapIndex points below the top.
−0.3 · no clear gap
Top
CoverageIndex weight that has results.
95% best
85%
SpeedOutput tokens per second.
81/s
—
EloLMArena rating and rank.
1455
—
ReleasedMonth of release.
2026-04
2026-10
Input / 1MUS dollars per 1M input tokens.
$0.65
—
Output / 1MUS dollars per 1M output tokens.
$3.41
—
1K chats2K tokens in, 500 out, each.
$3.01
—
1K documents20K tokens in, 1K out, each.
$16.41
—
1K agent tasks200K tokens in, 20K out, each.
$198.20
—
ContextMaximum tokens in one request.
262K
1M best
AgenticMulti-step tasks with tools.
59.5 best
55.0
CodingCode writing and repair.
59.3
60.8 best
ReasoningLogic problems and puzzles.
53.8
62.7 best
MultimodalTasks with images and text.
57.6
—
KnowledgeFacts and expert knowledge.
56.1
59.2 best
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
60.0
63.0 best
MathMath problems.
60.0 best
55.5

Benchmarks

10 shared
Only Kimi K2.6: 66Only Beam 1.0 501B-A23B NVFP4: 8
Kimi K2.6
Beam 1.0 501B-A23B NVFP4
81.0% best35 / 217
78.7%61 / 217
8.0%75 / 214
15.4% best48 / 214
80.2%19 / 114
80.7% best15 / 114
90.5% best26 / 84
89.3%36 / 84
58.6%38 / 83
65.6% best15 / 83
BrowseCompAgentic
83.2% best26 / 63
75.6%38 / 63
MCP AtlasAgentic
55.9%41 / 52
80.5% best15 / 52
52.2% best6 / 35
46.9%14 / 35
96.4%7 / 34
97.5% best4 / 34
92.5% best6 / 24
81.1%13 / 24