GPT-4.1 Mini

Apr 2025
GPT-4.1 Mini
Compare in
on
—per task?
$0.40 / $1.60per 1M in / outcache $0.10
158 tok/sfirst token in 1 s?
1Mcontext

Rankingin non-thinking mode?

Strong at

B
Tool calling#108 of 242
C
Agents & work tasks#122 of 175Reasoning#271 of 334Science#206 of 312Long context#187 of 273Knowledge#230 of 318Instructions#166 of 245Vision#97 of 124

Alternatives?

ModelScoreCostPer taskWhat changes
CheaperGPT OSS 120BhighOpen+29%-91%$0.092Context: 1M → 131K+1OpenR1 0528unstated×2.2same$0.39 / $1.63Context: 1M → 128K+1BetterGPT-5high×2.7+391%$1.25 / $10Context: 1M → 400KBettero3 Prohigh×2.6+4,900%$20 / $80Context: 1M → 200KBettero3high×2.5+400%$2 / $8Context: 1M → 200K

History

7 versions
0102030Oct 24Jan 25Apr 25Release dateAider Polyglot · Two attempts4o Mini4.1 · Apr
0102030Oct 24Jan 25Apr 25Release dateAider Polyglot · Two attempts4o Mini

Benchmarks

all categories
Chat & Knowledge
Humanity's Last Exam5#273
AA Omniscience Index-53.6#231
CritPt (physics)0#180
AA Intelligence Index 4.3.2≈10.2#206
AA Omniscience · Accuracy20.3#153
AA Omniscience · Hallucination Rate92.7#279
AA-LCR (long context) 1.144#189
GPQA Diamond66.4#208
AIME 2025 · AA archived46.3#91
Coding
Terminal-Bench Hard (version unstated)7.6#139
Terminal-Bench 2.110.1#127
LiveCodeBench · AA archived48.3#92
Agents & Tool Use
GDPval-AA (work tasks) 2.10#122
τ³-Banking · AA5.4#116
τ²-bench · AA52.9#110
Writing & Language
IFBench (instructions)38.3#167
COMET22 averagelab83.6
COMET22 multi-domain Zh-Enlab86.5
COMET22 multi-domain En-Zhlab82.6
COMET22 multi-domain En-Delab81.9
COMET22 multi-domain De-Enlab83.3
COMET22 WMT24lab84.6
Acceptance averagelab93
Acceptance En-Zhlab94
Acceptance Zh-Enlab98
Acceptance Zh-Xlab91.9
Acceptance X-Zhlab89.6
Excellence averagelab80
Excellence En-Zhlab80
Excellence Zh-Enlab90
Excellence Zh-Xlab77
Excellence X-Zhlab73.5
Vision & Documents
MMMU-Pro (vision)58.7#99

FAQ

How much does GPT-4.1 Mini cost?›

$0.40 per 1M input tokens and $1.60 per 1M output tokens; cached input $0.10 per 1M.

Is GPT-4.1 Mini free to use?›

There is no free API; the cheapest paid price is $0.40 per 1M input tokens.

How good is GPT-4.1 Mini, and what is it best at?›

#271 of 334 in Chat & Knowledge, #137 of 239 in Coding, #122 of 175 in Agents & Tool Use, #102 of 195 in Writing & Language, #97 of 124 in Vision & Documents.

Is GPT-4.1 Mini better than Qwen3 32B?›

On Aider Polyglot · two attempts, GPT-4.1 Mini scores 32.4 vs 40 — lower.

How can I use GPT-4.1 Mini?›

Through the API of Azure, Azure AI Foundry, OpenAI.

What is the context window of GPT-4.1 Mini?›

1M tokens, with up to 33K output tokens.

When was GPT-4.1 Mini released? Is there a newer version?›

Released 2025-04-14. It has been superseded by GPT-5 Mini.

Is GPT-4.1 Mini open source?›

No, GPT-4.1 Mini is a closed model available through an API. The closest open-weights option here is R1 0528.

Can GPT-4.1 Mini read images?›

It accepts text and images, PDF files as input.

What is GPT-4.1 Mini?›

GPT-4.1 Mini is a closed language model from OpenAI, released Apr 2025.

What are the best alternatives to GPT-4.1 Mini?›

GPT OSS 120B (cheaper), R1 0528 (open weights).