Models supported by GPT for Work extensions
Browse through the comprehensive list of models supported by each GPT for Work extension.
Without an API key
You can use the following models without an API key of your own.
* The cost indicated in the table applies if you're on self-service pricing. For more information, see GPT for Work pricing options.
** OpenAI o1 models split output tokens into completion tokens and reasoning tokens. Completion tokens are regular tokens visible to users, while reasoning tokens are internal tokens generated by the model for "thinking". The pricing includes both types of tokens. For more information about reasoning tokens and how they impact costs, see Reasoning models (OpenAI Platform).
With an API key
The following table shows the models/Assistants you can use if you create and set up your own API keys in GPT for Work extensions.
* The cost indicated in the table applies if you're on self-service pricing. For more information, see GPT for Work pricing options.
** OpenAI o1 models split output tokens into completion tokens and reasoning tokens. Completion tokens are regular tokens visible to users, while reasoning tokens are internal tokens generated by the model for "thinking". The pricing includes both types of tokens. For more information about reasoning tokens and how they impact costs, see Reasoning models (OpenAI Platform).
For a full list of models supported by Ollama, see the Ollama model library.
The tables on this page were created with Awesome Table Apps.