#how to adjust idleTimeoutSeconds

1 messages · Page 1 of 1 (latest)

gaunt sphinx
#

How do I adjust the idleTimeoutSeconds, I want to make it bigger because I keep getting this message:
"The model did not produce a response before the LLM idle timeout. Please try again, or increase agents.defaults.llm.idleTimeoutSeconds in your config (set to 0 to disable)."
when I use a big prompt

I am on version 2026.4.25-beta.11
this is ran on a mac mini, a computer dedicated to my openclaw only.
In the config I see this:
"agents": {
"defaults": {
"model": {
"primary": "minimax-portal/MiniMax-M2.7",
"fallbacks": []
},
"thinkingDefault": "low",
"compaction": {
"mode": "default"
},
"models": {
"minimax-portal/MiniMax-M2.7": {
"alias": "minimax-m2.7"
},
"minimax-portal/MiniMax-M2.7-highspeed": {
"alias": "minimax-m2.7-highspeed"
}
},
"workspace": "/Users/agentix/.openclaw/workspace",
"heartbeat": {
"every": "60s",
"model": "ollama/llama3.2:3b",
"lightContext": true
}
},

slim tartanBOT
#

To help others find answers, you can mark your question as solved via Right click solution message -> Apps -> ✅ Mark Solution

small graniteBOT
#

For 2026.4.25-beta.11, that key lives under agents.defaults.llm.

Fastest safe edit:

openclaw config set agents.defaults.llm.idleTimeoutSeconds 300 --strict-json
openclaw gateway restart

That sets it to 5 minutes. For really huge prompts you could use 600 (10 minutes). 0 disables the idle watchdog, but I’d only do that if you’re okay with a model request potentially hanging instead of failing cleanly.

If editing the config by hand, add llm as a sibling of heartbeat inside agents.defaults:

"heartbeat": {
  "every": "60s",
  "model": "ollama/llama3.2:3b",
  "lightContext": true