GPT-4.1
OpenAIGPTProprietary
Описание
GPT-4.1 is OpenAI's latest and most advanced flagship model, significantly improving upon GPT-4 Turbo in performance across benchmarks, speed, and cost-effectiveness.
Дата выхода
2025-04-14
Параметры
—
Длина контекста
1.0M
Модальности
image, pdf, text
Радар способностей
32
general
46
coding
49
reasoning
47
science
60
agents
85
multimodal
Рейтинги
| Домен | #Место | Оценка | Источник |
|---|---|---|---|
| Рейтинг кодинга | 309 | 49.0 | AA |
| Общий рейтинг | 303 | 43.0 | AA |
| Мультимодальный рейтинг | 66 | 54.0 | LS |
| Наука | 412 | 36.0 | AA |
Оценки бенчмарков (LLM Stats)
(LLM Stats (zeroeval))Chat
IFEvalGoogle Research (2023)
87.4%Сам.
Multi-IF
70.8%Сам.
TAU-bench Retail
68.0%Сам.
Multi-Challenge
38.3%Сам.
General
MMLU
90.2%Сам.
Aider-Polyglot Edit
52.9%Сам.
Aider-Polyglot
51.6%Сам.
Internal API instruction following (hard)
49.1%Сам.
Language
MMMLU
87.3%Сам.
COLLIE
65.8%Сам.
Long Context
ComplexFuncBench
65.5%Сам.
OpenAI-MRCR: 2 needle 128k
57.2%Сам.
OpenAI-MRCR: 2 needle 1M
46.3%Сам.
Math
MathVista
72.2%Сам.
AIME 2024
48.1%Сам.
AIME 2025
46.4%Сам.
HMMT 2025
28.9%Сам.
Multimodal
MMMU
74.8%Сам.
Reasoning
CharXiv-D
87.9%Сам.
GPQANYU + Cohere + Anthropic (2023)
66.3%Сам.
Graphwalks BFS <128k
61.7%Сам.
Graphwalks parents <128k
58.0%Сам.
CharXiv-R
56.7%Сам.
SWE-Bench Verified
54.6%Сам.
TAU-bench Airline
49.4%Сам.
Graphwalks parents >128k
25.0%Сам.
Graphwalks BFS >128k
19.0%Сам.
Humanity's Last Exam
5.4%Сам.
Vision
Video-MME (long, no subtitles)
72.0%Сам.
Индексы оценки AA
(Artificial Analysis)Math 500(OpenAI (2024), subset of Hendrycks et al. MATH (2021))91.3
Mmlu Pro(TIGER-Lab (Univ. of Waterloo, Toronto, CMU, 2024))80.6
Lcr(Artificial Analysis)68.3
Gpqa(NYU + Cohere + Anthropic (2023))66.6
Tau2(Sierra + U Toronto + Vector Institute (2025))47.1
Livecodebench(UC Berkeley + MIT + Cornell (2024))45.7
Aime(MAA (Mathematical Association of America))43.7
Ifbench(Google Research (2023))43.0
Math Index(Artificial Analysis)34.7
Aime 25(MAA (Mathematical Association of America))34.7
Terminalbench Hard(Stanford × Laude Institute (2026))13.6
Intelligence Index(Artificial Analysis)12.7
Hle(Center for AI Safety + Scale AI (2025))4.2
Оценки категорий LLM Stats
(LLM Stats (zeroeval))Legal90
Finance90
Instruction Following80
Language80
Healthcare80
Chat70
Multimodal70
Physics70
Structured Output70
Biology70
Chemistry70
Writing70
Reasoning60
General60
Communication60
Tool Calling60
Vision60
Math50
Frontend Development50
Code50
Long Context40
Spatial Reasoning40
Цены
Цена ввода$2 / 1M токенов
Цена вывода$8 / 1M токенов
Смешанная цена (3:1)$3.5 / 1M токенов
Цена чтения кэша$0.5 / 1M токенов
Скорость
Токенов/сек0.0
Задержка первого токена0.00s
Время до первого ответа0.00s
Рейтинг цен провайдеров
Рейтинг цен провайдеров
24 провайдеров
Самый дешевый: OpenAIСамый дорогой: Cortecs
ПровайдерВводВывод
1OpenAIСамый дешевый
$0
$0.00001
2Ofox
$1.6
$6.4
3Poe
$1.8
$7.2
4302.AI
$2
$8
5NanoGPT
$2
$8
6Abacus
$2
$8
7OpenRouter
$2
$8
8Kilo Gateway
$2
$8
9SAP AI Core
$2
$8
10Cloudflare AI Gateway
$2
$8
11Helicone
$2
$8
12Azure Cognitive Services
$2
$8
13Vercel AI Gateway
$2
$8
14DevPass (LLM Gateway)
$2
$8
15Azure
$2
$8
16FastRouter
$2
$8
17NEAR AI Cloud
$2
$8
18OrcaRouter
$2
$8
19Merge Gateway
$2
$8
20Pioneer
$2
$8
21Impossibl
$2
$8
22Eden AI
$2
$8
23LLM Gateway
$2
$8
24Cortecs
$2.192
$8.769
Сравнение цен разных API-провайдеров для этой модели.