4.11.0

4.11.0

• The model name parameters are now case sensitive.
  Previously, model names were converted to lowercase internally. This is optimal for the OpenAI, but some compatible API servers require uppercase model names.
• Add -All, -Limit & -Order params for Get-OpenAIFile.

4.10.0

4.10.0

• Add -Prediction param for Request-ChatCompletion.
  ref: Use Predicted Outputs

4.9.1

4.9.1

• Add a support for audio input / output in Request-ChatCompletion.

  Request-ChatCompletion -Model gpt-4o-audio-preview `
    -Modalities text, audio `
    -Voice 'shimmer' `
    -InputAudio 'C:\hello.mp3' `
    -AudioOutFile 'C:\response.mp3'

• Fix an issue with incorrect discontinuance date for the gpt4-32k model.

4.8.0

4.8.0

• Add an experimental support for Realtime API.
  This enables real-time conversations with the AI assitstant. Please refer to this guide.
  Guide: How to use Realtime API
• Update default api version of Azure OpenAI Service to 2024-10-01-preview
• Outputs Run and RunStep error messages as warnings.
• Fix an issue where some request URIs contains duplicate slashes.

4.6.0

4.6.0

• Add -Images parameter for Request-Moderation.

4.5.0

4.5.0

• Add omni-moderation-latest model to tab completions.

4.4.1

4.4.1

• Refactor file upload handling for non-ASCII filenames.

4.4.0

4.4.0

• Add o1-preview and o1-mini models to tab completions.
• Add -MaxCompletionTokens parameter for Request-ChatCompletion.
  The -MaxTokens parameter is now deprecated.
• gpt-3.5-turbo-0613 and gpt-3.5-turbo-16k-0613 is deprecated on 2024-09-13.
  You can still call these models, but it is removed from model name completion.

4.3.0

4.3.0

• Add -Include param for Get-ThreadRunStep.
• Add -RankerForFileSearch and -ScoreThresholdForFileSearch params for New-Assistant.
• Add -PollIntervalSec param for Wait-* functions.
• Update default api version of Azure OpenAI Service to 2024-07-01-preview
• Fix misc bugs related to Azure OpenAI Service.

4.2.0

4.2.0

• Add support for new chatgpt-4o-latest model.