Keimodel
Log inCompare
OpenAI

OpenAI

Frontier AI from the creators of ChatGPT

OpenAI is the research company behind ChatGPT, GPT-4, and the o-series reasoning models. Founded in 2015 with a focus on safe AGI development, OpenAI's models consistently lead on coding, instruction-following, and multimodal tasks. Their GPT-5 generation marks a significant capability leap, while the o-series excels at complex multi-step reasoning.

Founded 2015·8 models·Visit OpenAI

8 Models

GPT-5

OpenAI's flagship frontier model

5 cr
GPT-5 is OpenAI's most capable general-purpose model, combining deep reasoning, multimodal understanding, and long-context comprehension. It excels at complex analysis, creative tasks, and code generation.
128k ctxVisionTool use

GPT-5 Pro

Highest-capability GPT-5 variant

5 cr
GPT-5 Pro pushes the frontier further with extended compute budgets for reasoning and analysis. It is OpenAI's best model for the most demanding professional and research tasks.
128k ctxVisionTool use

GPT-5 Mini

Efficient GPT-5 for everyday tasks

2 cr
GPT-5 Mini delivers GPT-5-class reasoning in a smaller, faster, and more affordable package. It is ideal for high-volume applications where cost and latency matter.
128k ctxVisionTool use

GPT-5 Nano

Ultra-fast, ultra-affordable GPT-5

2 cr
GPT-5 Nano is OpenAI's smallest and fastest model, optimised for simple classification, extraction, and auto-complete tasks where sub-second latency is critical.
128k ctxTool useJSON mode

GPT-4o

Multimodal flagship — text, vision, and audio

3 cr
GPT-4o ('o' for omni) is OpenAI's multimodal model that processes text, images, and audio natively. It matches GPT-4 Turbo on text tasks while being twice as fast and half the price.
128k ctxVisionTool use

GPT-4o Mini

Affordable intelligence for high-volume tasks

2 cr
GPT-4o Mini is a lightweight version of GPT-4o that outperforms GPT-3.5 Turbo on most benchmarks while being 60% cheaper. It's the go-to model for building cost-effective AI products.
128k ctxVisionTool use

o3

OpenAI's most powerful reasoning model

5 cr
o3 uses extended chain-of-thought reasoning to tackle problems that require deep multi-step thinking: advanced mathematics, competitive programming, PhD-level science, and complex legal analysis.
200k ctxVisionTool use

o4 Mini

Fast, affordable reasoning for everyday problems

2 cr
o4 Mini brings the power of OpenAI's reasoning models to a more accessible price point. It performs exceptionally well on STEM tasks while responding significantly faster than o3.
200k ctxVisionTool use

Other providers

AnthropicAnthropicGoogleGoogleMetaMetaMistralMistral