Image processing fails when using GitHub CoPilot as LLM provider #5497
Description
Trying to reference an image file whenever using GitHub CoPilot as a LLM provider, fails with an error "Request failed: missing required Copilot-Vision-Request header for vision requests (code: ) (status 400)"
To Reproduce
Steps to reproduce the behavior:
- Go to codename goose, a desktop app running on Windows
- Select GoPilot->gpt-5-mini Model
- Go to new chat and upload an image file
- Ask goose to tell what is on the file
- See error above
Expected behavior
goose should be able to read the file and understand what is in it, whenever an LLM supporting vision is selected.
Please provide following information:
- OS & Arch: Windows Server 2022 Standard
- Interface: UI
- Version: 38.2.1
- Extensions enabled: official Atlassian, Azure DevOps, Brave MCPs, Developer Tools
- Provider & Model: CoPilot / gpt-mini-5
Additional context
If changing the provider to Openrouter and selecting the same model gpt-5-mini, the image is correctly processed.