OpenAI workflow block v4

[brunopicinin]

Description

New OpenAI block version (v4) to introduce breaking changes:

• Add GPT-5.1 and remove deprecated models
• Change default model to GPT-5.1
• Add object-detection task
• Add support for reasoning effort
• Migrate openai Chat Completions to Responses API
• Make max_tokens optional with model output limit as default
• Check for response status and improve error messages on error/fail/incomplete
• Fix temperature range validation (0.0-1.0) and prevent use with extended thinking
• Add metadata to show readable model names instead of IDs
• Update recommended parsers to v2 versions
• Add unit tests

Type of change

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)

How has this change been tested, please provide a testcase or example of how you tested the change?

Running inference locally, through workflows. Tested all task types, all models, and all reasoning efforts.

CleanShot.2025-12-04.at.01.16.16.mp4

Will the change affect Universe? If so was this change tested in universe?

No

[grzegorz-roboflow]

Compared with v3, LGTM

[brunopicinin]

Updated API proxy deployed to prod ✅