Frontier models are those that we are most likely to have come across - Claude, ChatGPT, Gemini. These are the highest performing models available. They are also known as closed source models. They are closed source because the training can cost hundreds and millions of dollars, and running them is extremely expensive due to their scale. There is often a charge to use the highest models.
An alternative to closed source models is open source models, although these are often more accurately described as open weight models, because whilst the model weights are available to the public, the underlying training data and specifics of the model are often not.
Frontier models, such as Claude, ChatGPT, Gemini, etc, often offer more than one model at any one time, optimised for different tasks or for economy. For example, at the time of writing Claude has Haiku, Sonnet and Opus models:
| Feature | Claude Opus 4.8 | Claude Sonnet 4.6 | Claude Haiku 4.5 |
|---|---|---|---|
| Description | Anthropic’s most capable model for complex reasoning and agentic coding | The best combination of speed and intelligence | The fastest model with near-frontier intelligence |
| Most frontier models support the OpenAI API format. |