[GH-ISSUE #16038] Support for images in tool responses #87897

Open
opened 2026-05-10 06:33:34 -05:00 by GiteaMirror · 0 comments
Owner

Originally created by @ZoranRavic on GitHub (May 7, 2026).
Original GitHub issue: https://github.com/ollama/ollama/issues/16038

Tool responses usually only contain text, but I've attempted to attach images (as base64, like we do in the user prompt) and apparently the vision models can not see the images in tool responses.

This would be a great feature for a tool such as "take_screenshot", which a model could use after calling other tool to confirm what happened.

Currently as a workaround we can inject a message with "role": "user" and attach images to it.

Originally created by @ZoranRavic on GitHub (May 7, 2026). Original GitHub issue: https://github.com/ollama/ollama/issues/16038 Tool responses usually only contain text, but I've attempted to attach images (as base64, like we do in the user prompt) and apparently the vision models can not see the images in tool responses. This would be a great feature for a tool such as "take_screenshot", which a model could use after calling other tool to confirm what happened. Currently as a workaround we can inject a message with `"role": "user"` and attach images to it.
GiteaMirror added the feature request label 2026-05-10 06:33:34 -05:00
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#87897