Inference
EN
OpenAI

GPT-4.1 mini

GPT-4.1 mini is a standard model from OpenAI in the GPT-4.1 family. 25 benchmarks count toward its index, in 7 categories.

  • Closed weights
  • Standard
  • Input: image · text · file
  • Released 2025-04
RankedEnough results for an index.
IndexAll categories, out of 100.34.5±3.7Median 51.0
CoverageIndex weight that has results.95%
SpeedOutput tokens per second.72/sMedian 67/s
First tokenSeconds to the first output token.0.9 sMedian 1.3 s
Input / 1MUS dollars per 1M input tokens.$0.40batch $0.20Median $0.60
Output / 1MUS dollars per 1M output tokens.$1.60batch $0.80 Batch price per 1M output tokens.Median $2.55
ContextMaximum tokens in one request.1.05MMedian 262K
EloLMArena rating and rank.1341±4 #216Median 1422

Details

From OpenRouter
API IDThe model ID on OpenRouter.
openai/gpt-4.1-mini
Max outputMaximum output tokens per request.
33K
Knowledge cutoffLast date of training data.
30 Jun 2024
ToolsTool calls on OpenRouter.
Yes
JSON outputOutput follows a given JSON schema.
Yes
Cache read / 1MUS dollars per 1M cached tokens.
$0.10
WeightsLink to the published weights.
—
RetiresDate OpenRouter removes the model.
—

Where to run

APIs that serve this model.
3 providers
ProviderTypeMaker, cloud, host or router.InputOutputContext
Microsoft AzureDeprecatedCloud$0.40$1.601.05M
OpenRouterRouter$0.40$1.601.05M
SAP AI CoreDeprecatedCloud$0.40$1.601.05M

US dollars per 1M tokens, list price. From models.dev (opens in a new tab) (MIT).

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.2 counted35.3188 / 218
CodingCode writing and repair.2 counted28.2199 / 209
ReasoningLogic problems and puzzles.6 counted36.8180 / 227
MultimodalTasks with images and text.1 counted39.2—
KnowledgeFacts and expert knowledge.6 counted36.0218 / 268
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.2 counted33.151 / 63
MathMath problems.6 counted36.5133 / 168

Results

25 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
MATH level 5Math87.3% ±0.720 / 88
Massive Multitask Language UnderstandingKnowledge87.5%10 / 34
Instruction-Following EvalInstruction88.5%31 / 70
MATH 500Math88.0% ±1.531 / 59
GPQA diamond · Epoch AIKnowledge65.8% ±2.7120 / 217
τ²-Bench Tool-Agent-User EvaluationAgentic52.9%93 / 149
OTIS Mock AIME 2024-2025Math44.7% ±6.1126 / 200
AIMEMath49.4% ±1.163 / 95
FrontierMath-2025-02-28-PrivateMath4.5% ±1.247 / 69
Critical Physics TasksReasoning0.0%150 / 214
Artificial Analysis GPQA DiamondKnowledge66.4%141 / 199
GPQA Diamond · Vals AIKnowledge67.9% ±2.499 / 135
Graduate-Level Google-Proof Q&AKnowledge64.2%105 / 141
GDPval-AA normalizedAgentic0.0%101 / 133
Artificial Analysis IFBenchInstruction38.3%111 / 146
Artificial Analysis Long Context ReasoningReasoning44.0%165 / 217
MMMU ProMultimodal70.5% ±1.170 / 92
SWE-bench Verified · swebench.comCoding23.9%35 / 46
LiveCodeBenchCoding58.2% ±1.2109 / 142
MMLU ProKnowledge77.2% ±0.4105 / 135
Artificial Analysis Omniscience AccuracyKnowledge20.3%128 / 164
Artificial Analysis Humanity's Last ExamKnowledge5.0%176 / 222
Artificial Analysis MMMU-ProMultimodal58.7%93 / 113
Mystery Game PuzzlesReasoning7.0% ±2.669 / 80
SWE-bench Verified · SWE-bench/experimentsCoding23.9%57 / 65
ARC-AGI-2 (semi-private)Reasoning0.0%120 / 131
SimpleQA VerifiedKnowledge12.7% ±1.185 / 91
τ²-bench TelecomAgentic48.9%17 / 18
FrontierMath-Tiers-1-3-v2-PrivateMath6.7% ±1.585 / 90
Software Engineering Benchmark VerifiedCoding23.6%111 / 114
ARC-AGI-1 (semi-private)Reasoning3.5%124 / 126
Chess PuzzlesReasoning1.0% ±1.089 / 90
τ²-bench RetailAgentic61.4%23 / 23
τ²-bench AirlineAgentic48.7%22 / 22
MortgageTaxFinance65.5% ±0.941 / 97
FrontierMath-2025-02-28-PublicMath10.0% ±10.029 / 64
TaxEval v2Finance71.9% ±0.970 / 143
MedQAHealthcare84.6% ±0.361 / 95
CorpFin v2Finance57.9% ±1.086 / 131
LegalBenchLegal78.0% ±0.4104 / 149
MGSMMultilingual87.8% ±1.957 / 74
ARC-AGI-2 (public eval)Reasoning0.0%102 / 122
ARC-AGI-1 (public eval)Reasoning7.2%113 / 116

Sources

BenchLMCC BY-NC 4.0OpenRouterTerms restrict reuseVals AINo licence statedEpoch AICC BYSierra ResearchMITSWE-bench teamNo licence statedARC Prize FoundationTerms restrict reusemodels.devMIT

History

5 releases
Index35.3 → 34.5
23 Oct 202612 Nov 2026
Output / 1M$1.60 → $1.60
23 Oct 202612 Nov 2026

Questions

How much does GPT-4.1 mini cost?
$0.40 per 1M input tokens, $1.60 per 1M output tokens.
How good is GPT-4.1 mini?
Its index is 34.5 out of 100. That is place 236 of 283 ranked models.
What is GPT-4.1 mini best at?
Math: 36.5 out of 100, place 133 of 168.
How fast is GPT-4.1 mini?
72 output tokens per second, 0.9 s to the first token.
What is the context window of GPT-4.1 mini?
1.05M tokens.
Is GPT-4.1 mini open source?
No. Its weights are not published.
When was GPT-4.1 mini released?
April 2025.

Badge

GPT-4.1 mini on Inference360: place 236 of 283, index 34.5
HTML
<a href="https://inference360.ai/models/gpt-4-1-mini"><img src="https://inference360.ai/badges/gpt-4-1-mini.svg" alt="GPT-4.1 mini on Inference360: place 236 of 283, index 34.5" height="20"></a>
Markdown
[![GPT-4.1 mini on Inference360: place 236 of 283, index 34.5](https://inference360.ai/badges/gpt-4-1-mini.svg)](https://inference360.ai/models/gpt-4-1-mini)

Compare with

Llama 4 Maverick34.4Mistral Large 334.3Gemini 2.5 Flash35.0Nemotron Ultra 253B35.9GPT-4.138.0

Lower price

dots3-note Preview63.4 · FreeApodex 1.1 Mini55.7 · FreeNemotron 3 Ultra50.8 · Free

More from OpenAI

All models
GPT-4.138.0GPT-4.1 nano26.3o351.8o4-mini44.7o4-mini (high)Unranked