openclaw - 💡(How to fix) Fix [Bug]: imageModel config not intercepting incoming message attachments as documented [1 participants]

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…
GitHub stats
openclaw/openclaw#72968Fetched 2026-04-28 06:29:18
View on GitHub
Comments
0
Participants
1
Timeline
2
Reactions
0
Author
Participants
Timeline (top)
labeled ×2

Attachments are dropped with "model does not support images" warnings despite a valid imageModel (Gemini 2.0 Flash) being configured to handle vision processing when the primary model (Tencent Hy3) lacks image support.

Error Message

No error response is sent to user — the image silently fails with no feedback

Root Cause

Gateway logs confirm imageModel is loaded: Image model: openrouter/google/gemini-2.0-flash-001:free User sends image via webchat → agents/tool-images resizes it (173.4KB → 66.9KB) Gateway drops attachment: parseMessageWithAttachments: 1 attachment(s) dropped — model does not support images The imageModel (Gemini 2.0 Flash) is never invoked to process the incoming attachment No image description is passed to the primary model (Tencent Hy3) User reports: "it disappears from the chat reply when you respond" — images show initially in UI but vanish on re-render because they weren't persisted to session history after being dropped No error response is sent to user — the image silently fails with no feedback

RAW_BUFFERClick to expand / collapse

Bug type

Behavior bug (incorrect output/state without crash)

Beta release blocker

No

Summary

Attachments are dropped with "model does not support images" warnings despite a valid imageModel (Gemini 2.0 Flash) being configured to handle vision processing when the primary model (Tencent Hy3) lacks image support.

Steps to reproduce

Configure OpenClaw with a text-only primary model: openclaw models set openrouter/tencent/hy3-preview:free Set imageModel to a vision-capable model: openclaw models set-image openrouter/google/gemini-2.0-flash-001:free Wait for config reload or restart gateway Send an image attachment via openclaw-control-ui (webchat) Check logs: tail -f /tmp/openclaw/openclaw-2026-04-27.log | grep "dropped|imageModel" Observe: parseMessageWithAttachments: 1 attachment(s) dropped — model does not support images appears, confirming the imageModel was not invoked

Expected behavior

According to https://docs.openclaw.ai/models:

agents.defaults.imageModel is used only when the primary model can't accept images.

When an image is sent, OpenClaw should:

Detect primary model (Tencent Hy3) lacks image support
Route the image to imageModel (Gemini 2.0 Flash) for processing
Pass the resulting image description to the primary model as context
Preserve the image in chat history (not disappear on re-render)

The imageModel should intercept incoming attachments and process them before the gateway drops them.

Actual behavior

Gateway logs confirm imageModel is loaded: Image model: openrouter/google/gemini-2.0-flash-001:free User sends image via webchat → agents/tool-images resizes it (173.4KB → 66.9KB) Gateway drops attachment: parseMessageWithAttachments: 1 attachment(s) dropped — model does not support images The imageModel (Gemini 2.0 Flash) is never invoked to process the incoming attachment No image description is passed to the primary model (Tencent Hy3) User reports: "it disappears from the chat reply when you respond" — images show initially in UI but vanish on re-render because they weren't persisted to session history after being dropped No error response is sent to user — the image silently fails with no feedback

OpenClaw version

2029.4.23

Operating system

windows 11

Install method

wsl ubuntu

Model

openrouter/tencent/hy3-preview:free

Provider / routing chain

openclaw openrouter

Additional provider/model setup details

No response

Logs, screenshots, and evidence

Impact and severity

No response

Additional information

No response

extent analysis

TL;DR

The issue can be fixed by ensuring the imageModel is correctly configured and invoked when the primary model lacks image support.

Guidance

  • Verify that the imageModel configuration is correctly set to openrouter/google/gemini-2.0-flash-001:free and that the primary model openrouter/tencent/hy3-preview:free is correctly detected as lacking image support.
  • Check the OpenClaw documentation to ensure that the agents.defaults.imageModel setting is correctly applied when the primary model cannot accept images.
  • Investigate why the imageModel is not being invoked to process the incoming attachment, despite being loaded and configured correctly.
  • Review the logs to see if there are any errors or warnings that may indicate why the image is being dropped and not persisted to the session history.

Example

No code snippet is provided as the issue seems to be related to configuration and model invocation rather than code.

Notes

The issue may be related to a misconfiguration or a bug in the OpenClaw model invocation logic. Further investigation is needed to determine the root cause.

Recommendation

Apply workaround: Verify the imageModel configuration and investigate why it's not being invoked, as the primary model lacks image support and the imageModel should be used in this case.

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…

FAQ

Expected behavior

According to https://docs.openclaw.ai/models:

agents.defaults.imageModel is used only when the primary model can't accept images.

When an image is sent, OpenClaw should:

Detect primary model (Tencent Hy3) lacks image support
Route the image to imageModel (Gemini 2.0 Flash) for processing
Pass the resulting image description to the primary model as context
Preserve the image in chat history (not disappear on re-render)

The imageModel should intercept incoming attachments and process them before the gateway drops them.

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING

openclaw - 💡(How to fix) Fix [Bug]: imageModel config not intercepting incoming message attachments as documented [1 participants]