Gemma 4 12B vs GPT-4.1

Change models
GPT-4.1OpenAI
IndexAll categories, out of 100.
38.3 ±3.9 best
38.0 ±3.0
GapIndex points below the top.
Top
−0.3 · no clear gap
CoverageIndex weight that has results.
100%
100%
SpeedOutput tokens per second.
—
81/s
EloLMArena rating and rank.
—
1383
ReleasedMonth of release.
2026-06
2025-04
Input / 1MUS dollars per 1M input tokens.
—
$2.00
Output / 1MUS dollars per 1M output tokens.
—
$8.00
Batch output / 1MBatch price per 1M output tokens.
—
$4.00
1K chats2K tokens in, 500 out, each.
—
$8.00
1K documents20K tokens in, 1K out, each.
—
$48.00
1K agent tasks200K tokens in, 20K out, each.
—
$560.00
ContextMaximum tokens in one request.
256K
1.05M best
AgenticMulti-step tasks with tools.
30.9
38.0 best
CodingCode writing and repair.
42.8 best
34.1
ReasoningLogic problems and puzzles.
41.2 best
40.4
MultimodalTasks with images and text.
35.4
41.9 best
KnowledgeFacts and expert knowledge.
39.5
41.2 best
MultilingualTasks in many languages.
30.0
45.0 best
InstructionTasks with strict prompt rules.
56.8 best
34.3
MathMath problems.
40.8 best
35.7

Benchmarks

17 shared
Only Gemma 4 12B: 12Only GPT-4.1: 42
Gemma 4 12B
GPT-4.1
EuroEval PolishMultilingual
43.1%110 / 227
54.1% best55 / 227
15.7% best122 / 216
4.2%188 / 216
EuroEval SwedishMultilingual
42.0%108 / 213
63.8% best25 / 213
63.7%132 / 211
68.3% best117 / 211
EuroEval GermanMultilingual
37.3%105 / 208
51.9% best33 / 208
0.0%145 / 207
0.0%145 / 207
45.5%95 / 199
54.4% best59 / 199
EuroEval FrenchMultilingual
42.8%121 / 195
62.3% best32 / 195
EuroEval SpanishMultilingual
32.3%120 / 194
52.8% best40 / 194
75.3% best109 / 193
66.6%134 / 193
EuroEval DutchMultilingual
44.9%94 / 192
62.6% best21 / 192
EuroEval ItalianMultilingual
41.5%105 / 191
57.1% best40 / 191
15.7%153 / 170
27.8% best90 / 170
78.8% best71 / 144
66.3%99 / 144
36.3%102 / 143
47.1% best92 / 143
73.5% best27 / 140
43.0%89 / 140
69.7% best73 / 109
61.2%86 / 109