GPT-5.4 mini vs Inkling

Change models
InklingThinking Machines Lab
IndexAll categories, out of 100.
54.5 ±2.4
54.8 ±2.9 best
GapIndex points below the top.
−0.3 · no clear gap
Top
CoverageIndex weight that has results.
100% best
95%
SpeedOutput tokens per second.
95/s best
85/s
EloLMArena rating and rank.
1412
1444 best
ReleasedMonth of release.
2026-03
2026-07
Input / 1MUS dollars per 1M input tokens.
$0.75 best
$1.00
Output / 1MUS dollars per 1M output tokens.
$4.50
$4.05 best
Batch output / 1MBatch price per 1M output tokens.
$2.25
—
1K chats2K tokens in, 500 out, each.
$3.75 best
$4.03
1K documents20K tokens in, 1K out, each.
$19.50 best
$24.05
1K agent tasks200K tokens in, 20K out, each.
$240.00 best
$281.00
ContextMaximum tokens in one request.
400K
524K best
AgenticMulti-step tasks with tools.
56.1 best
51.2
CodingCode writing and repair.
55.1 best
54.1
ReasoningLogic problems and puzzles.
52.2
56.1 best
MultimodalTasks with images and text.
54.8
55.0 best
KnowledgeFacts and expert knowledge.
53.7
57.1 best
MultilingualTasks in many languages.
57.6
—
InstructionTasks with strict prompt rules.
52.6
66.0 best
MathMath problems.
53.9
54.1 best

Benchmarks

47 shared
Only GPT-5.4 mini: 28Only Inkling: 31
GPT-5.4 mini
Inkling
28.1%91 / 222
31.9% best72 / 222
77.0%82 / 217
77.3% best77 / 217
86.9%65 / 217
88.3% best56 / 217
10.0% best65 / 214
5.4%81 / 214
88.9%63 / 200
88.9%63 / 200
87.5% best56 / 199
87.2%59 / 199
37.5%63 / 164
39.3% best56 / 164
TaxEval v2Finance
71.2%80 / 143
75.3% best19 / 143
81.5%60 / 142
85.5% best30 / 142
88.0% best38 / 141
87.9%39 / 141
83.1%57 / 135
87.1% best40 / 135
MMLU ProKnowledge
84.6%60 / 135
86.3% best43 / 135
25.6%71 / 133
28.6% best65 / 133
18.9%61 / 131
36.5% best54 / 131
CorpFin v2Finance
60.9%62 / 131
68.6% best7 / 131
63.7%64 / 126
79.5% best54 / 126
48.0% best70 / 123
19.2%94 / 123
17.8%55 / 122
37.8% best48 / 122
75.1%55 / 116
90.6% best44 / 116
52.1% best42 / 114
47.0%66 / 114