v2.1.0

What's new in v2.1.0

GPT-5 models now work, and the setup fields are properly labelled.

Fixed

• GPT-5 / reasoning models (gpt-5, gpt-5-mini, gpt-5.4-mini, gpt-5.5, …) now work for chat/data tasks — they previously failed with max_tokens then temperature 400 errors. (#29)
• Config and Options fields now show readable labels (e.g. Chat Model) instead of raw keys like chat_model.

Changed

• Chat API version bumped to 2025-04-01-preview (needed for GPT-5 / max_completion_tokens).
• Reasoning models get a larger completion budget and a clear error if they return no content.

Note

Enter your Azure AI Foundry deployment name as the Chat Model (e.g. gpt-5-mini). Requires a GPT-5 deployment.

Full Changelog: v2.0.4...v2.1.0