GPT-5.4 mini vs GPT-5.4 nano

Change models
IndexAll categories, out of 100.
54.5 ±2.4 best
50.9 ±2.5
GapIndex points below the top.
Top
−3.6
CoverageIndex weight that has results.
100%
100%
SpeedOutput tokens per second.
95/s best
91/s
EloLMArena rating and rank.
1412 best
1372
ReleasedMonth of release.
2026-03
2026-03
Input / 1MUS dollars per 1M input tokens.
$0.75
$0.20 best
Output / 1MUS dollars per 1M output tokens.
$4.50
$1.25 best
Batch output / 1MBatch price per 1M output tokens.
$2.25
$0.625 best
1K chats2K tokens in, 500 out, each.
$3.75
$1.03 best
1K documents20K tokens in, 1K out, each.
$19.50
$5.25 best
1K agent tasks200K tokens in, 20K out, each.
$240.00
$65.00 best
ContextMaximum tokens in one request.
400K
400K
AgenticMulti-step tasks with tools.
56.1 best
50.5
CodingCode writing and repair.
55.1 best
53.7
ReasoningLogic problems and puzzles.
52.2
53.7 best
MultimodalTasks with images and text.
54.8 best
41.7
KnowledgeFacts and expert knowledge.
53.7 best
44.5
MultilingualTasks in many languages.
57.6 best
44.4
InstructionTasks with strict prompt rules.
52.6
61.3 best
MathMath problems.
53.9
55.9 best

Benchmarks

69 shared
Only GPT-5.4 mini: 6Only GPT-5.4 nano: 5
GPT-5.4 mini
GPT-5.4 nano
28.1%91 / 222
28.3% best90 / 222
77.0% best82 / 217
76.7%84 / 217
86.9% best65 / 217
78.5%94 / 217
10.0% best65 / 214
9.3%69 / 214
EuroEval PolishMultilingual
62.5% best13 / 209
55.2%41 / 209
88.9% best63 / 200
87.8%68 / 200
87.5% best56 / 199
81.7%93 / 199
EuroEval SwedishMultilingual
68.9% best12 / 195
59.1%36 / 195
EuroEval GermanMultilingual
57.8% best12 / 191
49.2%42 / 191
64.7% best19 / 182
56.7%42 / 182
EuroEval FrenchMultilingual
67.4% best16 / 178
54.9%64 / 178
EuroEval SpanishMultilingual
62.8% best9 / 176
53.4%33 / 176
EuroEval DutchMultilingual
63.1% best16 / 175
53.1%55 / 175
EuroEval ItalianMultilingual
66.2% best14 / 174
55.7%45 / 174
37.5% best63 / 164
25.7%102 / 164
93.4% best31 / 149
92.5%34 / 149
73.3%29 / 146
75.9% best16 / 146
TaxEval v2Finance
71.2% best80 / 143
67.4%107 / 143
81.5%60 / 142
84.0% best41 / 142
88.0% best38 / 141
82.8%61 / 141