Granite 4.2 8B vs MiniCPM5-2B

Change models
MiniCPM5-2BOpenBMB
IndexAll categories, out of 100.
37.6 ±5.4 best
37.1 ±5.0
GapIndex points below the top.
Top
−0.5 · no clear gap
CoverageIndex weight that has results.
85%
85%
SpeedOutput tokens per second.
101/s
—
EloLMArena rating and rank.
1319
—
ReleasedMonth of release.
2026-08
2026-09
Input / 1MUS dollars per 1M input tokens.
$0.06
—
Output / 1MUS dollars per 1M output tokens.
$0.15
—
1K chats2K tokens in, 500 out, each.
$0.195
—
1K documents20K tokens in, 1K out, each.
$1.35
—
1K agent tasks200K tokens in, 20K out, each.
$15.00
—
ContextMaximum tokens in one request.
131K
131K
AgenticMulti-step tasks with tools.
36.8
38.2 best
CodingCode writing and repair.
36.5 best
32.0
ReasoningLogic problems and puzzles.
37.8
39.1 best
MultimodalTasks with images and text.
—
—
KnowledgeFacts and expert knowledge.
33.3
34.7 best
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
63.2 best
44.4
MathMath problems.
41.9
43.9 best

Benchmarks

15 shared
Only Granite 4.2 8B: 5Only MiniCPM5-2B: 9
Granite 4.2 8B
MiniCPM5-2B
9.7% best148 / 222
8.9%151 / 222
45.0%164 / 217
59.0% best143 / 217
0.3%137 / 214
0.3%137 / 214
63.1%150 / 199
70.2% best132 / 199
0.0%101 / 133
10.4% best92 / 133
31.5% best104 / 114
26.3%106 / 114
47.7% best101 / 114
46.4%104 / 114
20.6% best83 / 86
8.6%86 / 86
19.1% best82 / 83
14.4%83 / 83
74.0% best53 / 75
70.8%55 / 75
73.2% best38 / 59
69.1%42 / 59
79.3% best16 / 57
66.3%42 / 57
86.7% best24 / 53
86.5%25 / 53
36.1% best27 / 35
26.3%34 / 35
52.4%16 / 26
66.6% best11 / 26