Цены на LLM API

Модели OpenAI

Все модели OpenAI, доступные в TokenHot: 17.

Один API-ключ открывает доступ к 17 моделям OpenAI с актуальными ценами и без отдельной учётной записи провайдера.

OpenAI
Контекст1.05M
Вход
$0.4000/M$2.0000/M
Вывод
$2.0000/M$10.0000/M
The sensible default for most work: code review, multistep agent workflows, everyday writing such as turning meeting notes into a project brief, and batch processing. Priced the same as GPT-6.1 Sol but smarter and more token-efficient, so a plan allowance goes further; it lands close to flagship Astra on coding benchmarks at about a fifth of the cost — start here and move to Astra only for genuinely hard problems.
Тип входа:
Тип выхода:
ReasoningTool UseFunction CallingStructured OutputLong ContextWeb SearchCode Execution
Попробовать модель
OpenAI
Контекст1.05M
Вход
$1.6000/M$8.0000/M
Вывод
$8.0000/M$40.0000/M
Reserve this for the hardest work: tracing a bug across services until tests pass, producing slides, spreadsheets, or legal documents that hold up under review, cross-checking a research question across sources, or CAD-precise computer use in 3D and game projects. As the flagship tier it costs $40/M output tokens; for routine coding and writing, sibling Sol gets within half a point on code-merge benchmarks at roughly a fifth of the cost.
Тип входа:
Тип выхода:
ReasoningTool UseFunction CallingStructured OutputLong ContextWeb SearchCode Execution
Попробовать модель
OpenAI
Контекст1.05M
Вход$1.0000/M
Вывод$5.0000/M
A good fit for well-scoped everyday work that involves images, computer interfaces, or a clear outcome. Its strengths are visual reasoning and precise computer use, with clear, direct writing; it is faster and lower-cost than GPT-6.1 Sol, while the higher tier remains preferable for the hardest reasoning and coding.
Тип входа:
Тип выхода:
ReasoningTool UseFunction CallingStructured OutputLong ContextWeb SearchCode Execution
Попробовать модель
OpenAI
Контекст1.05M
Вход
$0.4000/M$2.0000/M
Вывод
$2.0000/M$10.0000/M
For complex coding, computer use, and professional document work. It improves on GPT-6 Sol in demanding tasks and approaches Astra at a lower cost; choose Astra when maximum quality is the priority.
Тип входа:
Тип выхода:
ReasoningWeb SearchTool UseFunction CallingStructured OutputLong ContextCode Execution
Попробовать модель
OpenAI
OpenAI
Контекст1.05M
Вход
$0.4000/M$2.0000/M
Вывод
$2.0000/M$10.0000/M
Built for complex coding and agentic workflows, including project-level programming, long-horizon planning, and multi-step tool use. Its 1.05M-token context, text/image input, reasoning, function calling, and structured outputs make it the workhorse choice; it costs more than GPT-6 Luna when quality and execution stability matter most.
Тип входа:
Тип выхода:
ReasoningWeb SearchTool UseFunction CallingStructured OutputLong ContextCode Execution
Попробовать модель
OpenAI
Контекст1.05M
Вход
$0.0200/M$0.1000/M
Вывод
$0.1000/M$0.5000/M
Designed for focused, high-volume work such as batch summarization, classification, rewriting, and frequent agent calls. It still offers a 1.05M-token context, text/image input, reasoning, function calling, and structured outputs, but trades depth for much lower cost; prefer GPT-6 Sol for the hardest long-horizon tasks.
Тип входа:
Тип выхода:
ReasoningWeb SearchTool UseFunction CallingStructured OutputLong ContextCode Execution
Попробовать модель
Контекст400K
ЗА ИЗОБР.$0.0080/image
The discounted-price route for GPT Image 2.5 Sunburst, mapped by the current channel to Sunburst. Sale indicates a special-price route and does not change Sunbursts role in high-precision image generation and editing. Best for cost-sensitive professional visual work that still needs reference-image editing and detail control; use regular gpt-image-2.5-sunburst when you want the standard route.
Тип входа:
Тип выхода:
Multimodal OutputLong Context
Попробовать модель
Контекст400K
ЗА ИЗОБР.$0.0480/image
Built for premium visual workflows where editing precision and finished-image quality matter most: a better fit for production campaign creative, polished product imagery, and multi-turn edits that need to preserve the subject, composition, and visual treatment. It accepts text and reference-image inputs and produces images, with longer generation times than Flare.
Тип входа:
Тип выхода:
Multimodal OutputLong Context
Попробовать модель
Контекст400K
ЗА ИЗОБР.$0.0080/image
The discounted-price route for GPT Image 2.5 Flare, mapped by the current channel to Flare. Sale indicates a special-price route and does not change Flares role in fast, high-quality everyday image generation and editing. Best for cost-sensitive, frequent image tasks; use regular gpt-image-2.5-flare when you want the standard route.
Тип входа:
Тип выхода:
Multimodal OutputLong Context
Попробовать модель
Контекст400K
ЗА ИЗОБР.$0.0480/image
A strong default for high-quality image generation and editing: optimized for faster iteration across creator and social content, product experiences, visual search, rapid prototyping, and high-volume generation. It accepts text and reference-image inputs and produces images; choose Sunburst when tighter editing control matters more than speed.
Тип входа:
Тип выхода:
Multimodal OutputLong Context
Попробовать модель
OpenAI
Контекст1.05M
Вход
$2.0000/M$10.0000/M
Вывод
$10.0000/M$50.0000/M
A flagship end-to-end model for complex reasoning, software engineering, computer use, deep research, and document work. It offers about a 1.05M-token context window, accepts text, images, and documents, returns text, and supports reasoning, search, function calling, structured outputs, code execution, and tool orchestration. Choose it over lower-tier models for long-horizon task quality and execution stability, with higher usage cost as the trade-off.
Тип входа:
Тип выхода:
ReasoningWeb SearchTool UseFunction CallingStructured OutputLong ContextCode Execution
Попробовать модель
OpenAI
Контекст1.05M
Вход
$0.4000/M$2.0000/M
Вывод
$2.4000/M$12.0000/M
The balanced workhorse tier of GPT-5.6, combining capability and cost. Well suited to production work involving image understanding, tool use, and large source material; choose Sol for the hardest quality-first work or Luna for cost-controlled high volume.
Тип входа:
Тип выхода:
ReasoningFunction CallingStructured OutputLong Context
Попробовать модель
OpenAI
Контекст1.05M
Вход$5.0000/M
Вывод$30.0000/M
The flagship GPT-5.6 tier for complex professional work and quality-first delivery. It supports image input, function calling, structured outputs, and an exceptionally large context window; consider Terra or Luna when latency or unit cost is the primary constraint.
Тип входа:
Тип выхода:
ReasoningFunction CallingStructured OutputLong Context
Попробовать модель
OpenAI
Контекст1.05M
Вход$0.2000/M
Вывод$1.2000/M
The cost-sensitive, high-volume tier of GPT-5.6. A good fit for summarization, classification, rewriting, and batch automation; choose the Sol sibling for complex professional judgment or quality-first multi-step work.
Тип входа:
Тип выхода:
ReasoningFunction CallingStructured OutputLong Context
Попробовать модель
OpenAI
Контекст400K
ЗА ИЗОБР.$0.0080/image
The discounted-price route for GPT Image 2, mapped by the current channel to the same underlying model. Sale indicates a special-price route, not a separate capability tier. A good fit for cost-sensitive everyday generation and reference-image editing; choose regular gpt-image-2 when route consistency matters more.
Тип входа:
Тип выхода:
Multimodal OutputLong Context
Попробовать модель
OpenAI
Контекст128K
ЗА ИЗОБР.$0.0280/image
Создан для высококачественного создания и редактирования изображений, улучшен по сравнению с GPT Image 1.x в рендеринге текста, локальном редактировании и многошаговой согласованности. По сравнению с моделями изображений Gemini, Qwen Image и Seedream, он хорошо подходит для визуальных эффектов бренда, дизайна рекламы, изображений продуктов и строго контролируемых творческих рабочих процессов.
Тип входа:
Тип выхода:
OpenAI
Вход
$15.0000/M$75.0000/M
Вывод
$15.0000/M$75.0000/M