GLM-5-Turbo vs Kimi K2.5

Change models
Kimi K2.5Moonshot AI
IndexAll categories, out of 100.
51.7 ±10.2 best
51.6 ±2.2
GapIndex points below the top.
Top
−0.1 · no clear gap
CoverageIndex weight that has results.
55%
100% best
SpeedOutput tokens per second.
42/s best
22/s
EloLMArena rating and rank.
—
—
ReleasedMonth of release.
2026-03
2026-01
Input / 1MUS dollars per 1M input tokens.
$1.20
$0.50 best
Output / 1MUS dollars per 1M output tokens.
$4.00
$2.50 best
1K chats2K tokens in, 500 out, each.
$4.40
$2.25 best
1K documents20K tokens in, 1K out, each.
$28.00
$12.50 best
1K agent tasks200K tokens in, 20K out, each.
$320.00
$150.00 best
ContextMaximum tokens in one request.
203K
262K best
AgenticMulti-step tasks with tools.
56.9 best
48.9
CodingCode writing and repair.
—
53.2
ReasoningLogic problems and puzzles.
46.5
51.1 best
MultimodalTasks with images and text.
—
51.6
KnowledgeFacts and expert knowledge.
51.2
53.2 best
MultilingualTasks in many languages.
—
52.1
InstructionTasks with strict prompt rules.
56.6 best
52.8
MathMath problems.
—
53.1

Benchmarks

8 shared
Only Kimi K2.5: 81
GLM-5-Turbo
Kimi K2.5
27.8%89 / 216
30.7% best73 / 216
71.7%101 / 211
78.0% best70 / 211
0.3%132 / 207
3.1% best93 / 207
84.7%75 / 193
87.9% best51 / 193
28.4%84 / 170
35.2% best68 / 170
98.5% best3 / 143
95.9%13 / 143
73.2% best30 / 140
70.2%42 / 140
Claw-EvalAgentic
55.8% best35 / 42
52.3%37 / 42