OpenAI
State-of-the-art AI models for every application
Last reviewed Mar 14, 2026
OpenAI is an AI research company offering its GPT model family, reasoning models, image generation, and realtime audio models through its API platform. Developers build on the Responses API and Agents SDK with pay-per-token pricing.
LLM API Pricing
Pay-per-token pricing. Prices shown per 1M tokens.
Prices last updated: October 6, 2026
| Model | Input/1M | Output/1M | |||
|---|---|---|---|---|---|
| $0.050 | $0.400 | ||||
| $0.100 | $0.400 | ||||
| $0.150 | $0.600 | ||||
| $0.200 | $1.25 | ||||
| $0.200 | $1.20 | ||||
| $0.250 | $2.00 | ||||
| $0.400 | $1.60 | ||||
| $0.500 | $1.50 | ||||
| $0.750 | $4.50 | ||||
| $1.10 | $4.40 | ||||
OpenAI pricing by model
Input, output, and batch rates, plus alternatives, for one model at a time.
Pros & Cons
Advantages
- Industry-leading model performance and accuracy
- Extensive training data providing unmatched breadth across languages and domains
- Developer-friendly SDKs for Python, JavaScript, and other languages
- Comprehensive documentation and large community
- Multimodal capabilities across text, audio, image, and video
- Strong enterprise features with data residency options
Limitations
- Higher cost compared to open-source alternatives
- Token usage can scale quickly with conversation threads
- Data privacy concerns for sensitive applications
- Closed-source models with limited transparency
- Rate limits can be restrictive for high-volume applications
Key Features
GPT Model Family
GPT-6 Astra for complex reasoning and coding, GPT-6 Sol to balance capability and cost, and GPT-6 Luna for cost-sensitive, high-volume workloads
Reasoning Models
Reasoning models for multi-step, complex tasks
Image Generation
Generate images as output with GPT Image 2
Realtime and Audio
Realtime API for voice applications plus endpoints to analyze, transcribe, and generate audio
Multimodal Capabilities
Process text, images, audio, and video inputs with unified models
Function Calling
Structured outputs and tool use for building agents and workflows
Responses API and Agents SDK
Direct model requests with tools and multimodal input via the Responses API; code-first agents with tools, handoffs, approvals, and tracing via the Agents SDK
Fine-Tuning and Evals
Adapt models to specific use cases with fine-tuning, evals, and distillation
Pricing Options
| Option | Details |
|---|---|
| Pay-per-token | Charged based on tokens processed for both input and output |
| Batch API | 50% discount for asynchronous batch processing |
| Prompt Caching | Discounts for recently seen input tokens |
Availability & Support
Regions
Global availability with data residency options in US, Europe, UK, Canada, Japan, South Korea, Singapore, India, Australia, and UAE
Support
Documentation, Help Center, Developer Forum, email support, and enterprise support plans with dedicated account managers
Getting Started
- 1
Create an account
Sign up at platform.openai.com
- 2
Generate API key
Create an API key in the dashboard and store it securely
- 3
Install SDK
Install the OpenAI SDK for Python, JavaScript, or your preferred language
- 4
Make your first call
Use the SDK to call OpenAI models for text generation or other tasks
Frequently Asked Questions
Which models does OpenAI serve?
Check the pricing table above for the models OpenAI currently serves and their per-token rates.
How do I get started with OpenAI?
Create an account, Generate API key, Install SDK, Make your first call
What are OpenAI's main advantages?
OpenAI's main advantages include: Industry-leading model performance and accuracy, Extensive training data providing unmatched breadth across languages and domains, Developer-friendly SDKs for Python, JavaScript, and other languages, Comprehensive documentation and large community, Multimodal capabilities across text, audio, image, and video, Strong enterprise features with data residency options.
What are OpenAI's limitations?
OpenAI's main limitations include: Higher cost compared to open-source alternatives, Token usage can scale quickly with conversation threads, Data privacy concerns for sensitive applications, Closed-source models with limited transparency, Rate limits can be restrictive for high-volume applications.
Compare Providers
Find the best prices for the same GPUs and models from other providers