Gemini 3.1 Flash-Lite vs Kimi K2.5

Change models
Kimi K2.5Moonshot AI
IndexAll categories, out of 100.
52.0 ±8.4
52.0 ±2.2
GapIndex points below the top.
Top
Top
CoverageIndex weight that has results.
70%
100% best
SpeedOutput tokens per second.
128/s best
23/s
EloLMArena rating and rank.
—
—
ReleasedMonth of release.
2026-03
2026-01
Input / 1MUS dollars per 1M input tokens.
$0.25 best
$0.50
Output / 1MUS dollars per 1M output tokens.
$1.50 best
$2.50
1K chats2K tokens in, 500 out, each.
$1.25 best
$2.25
1K documents20K tokens in, 1K out, each.
$6.50 best
$12.50
1K agent tasks200K tokens in, 20K out, each.
$80.00 best
$150.00
ContextMaximum tokens in one request.
1.05M best
262K
AgenticMulti-step tasks with tools.
46.4
49.3 best
CodingCode writing and repair.
—
53.6
ReasoningLogic problems and puzzles.
59.9 best
51.4
MultimodalTasks with images and text.
51.2
52.2 best
KnowledgeFacts and expert knowledge.
52.0
53.5 best
MultilingualTasks in many languages.
—
52.5
InstructionTasks with strict prompt rules.
—
53.3
MathMath problems.
50.2
53.5 best

Benchmarks

4 shared
Only Gemini 3.1 Flash-Lite: 5Only Kimi K2.5: 85
Gemini 3.1 Flash-Lite
Kimi K2.5
81.8%82 / 217
87.6% best62 / 217
80.0%86 / 200
92.2% best54 / 200
38.5%41 / 52
45.9% best29 / 52
73.2%47 / 50
77.5% best41 / 50