Issue Station
Sign In
Tools hereToolbox

openclaw - 💡(How to fix) Fix [Bug]: Gateway default model not updated after `openclaw models set` + service restart [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#57939•Fetched 2026-04-08 01:55:57
View on GitHub
Comments
0
Participants
1
Timeline
1
Reactions
0
Author
shunterashuntera
Participants
shunterashuntera
Timeline (top)
labeled ×1

After setting a new default model with openclaw models set and performing a full gateway stop/start, new Telegram sessions continue to report the old default model rather than the newly configured one.

Root Cause

After setting a new default model with openclaw models set and performing a full gateway stop/start, new Telegram sessions continue to report the old default model rather than the newly configured one.

RAW_BUFFERClick to expand / collapse

Bug type

Behavior bug (incorrect output/state without crash)

Beta release blocker

No

Summary

After setting a new default model with openclaw models set and performing a full gateway stop/start, new Telegram sessions continue to report the old default model rather than the newly configured one.

Steps to reproduce

  1. Run openclaw models set ollama/nemotron-3-super:cloud
  2. Verify with openclaw config get agents.defaults.model — confirms new default is set
  3. Run systemctl --user stop openclaw-gateway.service && sleep 3 && systemctl --user start openclaw-gateway.service
  4. Send /new in Telegram
  5. Observe session confirmation message reports old default (openai-codex/gpt-5.4) not the new one

Expected behavior

New sessions started after a gateway restart should use the default model set by openclaw models set.

From Telegram transcript:

[bot]: Hey user. Monday afternoon, what's on your mind? [bot]: ✅ New session started · model: anthropic/claude-opus-4-6 (default: openai-codex/gpt-5.4) [bot]: Hey user. Monday afternoon, what's on your mind? (Running on Opus right now — default's set to GPT-5.4 if you want to switch.) user: /new [bot]: ✅ New session started · model: anthropic/claude-opus-4-6 (default: openai-codex/gpt-5.4) [bot]: Hey user. Monday afternoon, what are we getting into?

Actual behavior

New sessions report openai-codex/gpt-5.4 as the default despite openclaw config get agents.defaults.model and openclaw models status both confirming ollama/nemotron-3-super:cloud as the primary default.

OpenClaw version

OpenClaw 2026.3.28 (f9b1079)

Operating system

Ubuntu 24.04.4 LTS

Install method

npm via nvm (Node v22.22.0)

Model

ollama/nemotron-3-super:cloud

Provider / routing chain

ollama (cloud-routed)

Additional provider/model setup details

No response

Logs, screenshots, and evidence

Impact and severity

No response

Additional information

No response

extent analysis

Fix Plan

To resolve the issue of new Telegram sessions reporting the old default model, follow these steps:

  • Update the gateway configuration: Ensure that the openclaw-gateway.service is properly configured to reload the latest model settings after a restart.
  • Verify model configuration: Confirm that the openclaw models set command has successfully updated the default model in the configuration file.
  • Clear cache: Clear any cached model settings that may be causing the old default model to be used.

Example code to update the gateway configuration:

# Update the gateway configuration to reload model settings
sudo systemctl --user edit openclaw-gateway.service

Add the following lines to the [Service] section:

[Service]
Restart=always
Environment=OPENCLAW_MODELS_RELOAD=true

Example code to verify model configuration:

# Verify the default model setting
openclaw config get agents.defaults.model

This should output the newly set default model ollama/nemotron-3-super:cloud.

Example code to clear cache:

# Clear cached model settings
openclaw models clear-cache

Verification

After completing these steps, restart the openclaw-gateway.service and send a new session request in Telegram to verify that the correct default model is being used:

# Restart the gateway service
systemctl --user stop openclaw-gateway.service && sleep 3 && systemctl --user start openclaw-gateway.service

# Send a new session request in Telegram
/user /new

The response should indicate that the new default model ollama/nemotron-3-super:cloud is being used.

Extra Tips

  • Ensure that the openclaw-gateway.service is properly configured to reload the latest model settings after a restart.
  • Regularly clear cached model settings to prevent outdated models from being used.
  • Verify that the openclaw models set command has successfully updated the default model in the configuration file.

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

New sessions started after a gateway restart should use the default model set by openclaw models set.

From Telegram transcript:

[bot]: Hey user. Monday afternoon, what's on your mind? [bot]: ✅ New session started · model: anthropic/claude-opus-4-6 (default: openai-codex/gpt-5.4) [bot]: Hey user. Monday afternoon, what's on your mind? (Running on Opus right now — default's set to GPT-5.4 if you want to switch.) user: /new [bot]: ✅ New session started · model: anthropic/claude-opus-4-6 (default: openai-codex/gpt-5.4) [bot]: Hey user. Monday afternoon, what are we getting into?

#API middleware#SSR setup#ISR setup#authentication setup#request error

Still need to ship something?

×6

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

Back to top recommendations

TRENDING