timeoutSeconds | Friends of the Crustacean 🦞🤝 | Page 1

slender wagon May 13, 2026, 4:21 PM

#

Hey krill I am getting this error any solution. Using ai studio api with gemma4 31b. Openclaw version 2026.5.7

Error:
The model did not produce a response before the model idle timeout. Please try again, or increase models.providers.<id>.timeoutSeconds for slow local or self-hosted providers.

torn cedarBOT May 13, 2026, 4:21 PM

#

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

wheat groveBOT May 13, 2026, 4:23 PM

#

Yep — that error means OpenClaw started the model request, but AI Studio/Gemma didn’t send any output before OpenClaw’s idle watchdog fired.

Try increasing the timeout for the provider id you’re using. The <id> is the provider part of your model string / config entry, e.g. if your model is google/gemma..., the id is likely google; if you made a custom provider like aistudio, use that.

openclaw config set models.providers.<id>.timeoutSeconds 300
openclaw gateway restart

If it still times out, try 600.

On some builds the relevant global idle watchdog is:

openclaw config set agents.defaults.llm.idleTimeoutSeconds 300
openclaw gateway restart

#

I’d also run this to confirm the provider/model is reachable and not failing upstream:

openclaw models status
openclaw logs --follow

Gemma 31B via AI Studio can be slow to first-token, so this is usually just “wait longer,” not necessarily a broken setup. Docs refs: https://docs.openclaw.ai/help/faq and https://docs.openclaw.ai/gateway/configuration

slender wagon May 13, 2026, 4:26 PM

#

Any idea on what to replace id as:
Config : ~/.openclaw/openclaw.json
Agent dir : ~/.openclaw/agents/main/agent
Default : google/gemma-4-31b-it
Fallbacks (0) : -
Image model : -
Image fallbacks (0): -
Aliases (1) : gemini -> google/gemini-3.1-pro-preview
Configured models (2): google/gemini-3.1-pro-preview, google/gemma-4-31b-it

wheat groveBOT May 13, 2026, 4:26 PM

#

Use google there — it’s the provider id from google/gemma-4-31b-it.

So try:

openclaw config set models.providers.google.timeoutSeconds 300
openclaw gateway restart

If that config key doesn’t exist / doesn’t change the behavior on 2026.5.7, use the LLM idle watchdog setting instead:

openclaw config set agents.defaults.llm.idleTimeoutSeconds 300
openclaw gateway restart

For Gemma 31B on AI Studio, I’d start with 300; if it still fails before first output, bump to 600.

slender wagon May 13, 2026, 4:27 PM

#

This if how much config looks like
"agents": {
"defaults": {
"workspace": "/home/beryl-server/.openclaw/workspace",
"model": {
"primary": "google/gemma-4-31b-it"
},
"models": {
"google/gemini-3.1-pro-preview": {},
"google/gemma-4-31b-it": {}
},
"timeoutSeconds": 18000
}
},
"session": {
"dmScope": "per-channel-peer"
},

wheat groveBOT May 13, 2026, 4:27 PM

#

Yep — in your config, agents.defaults.timeoutSeconds: 18000 is the overall agent/run timeout. It does not necessarily stop the “model produced no chunks yet” idle timeout.

Your provider id is still google, so add a top-level models.providers.google block, as a sibling of agents and session:

{
  "agents": {
    "defaults": {
      "workspace": "/home/beryl-server/.openclaw/workspace",
      "model": {
        "primary": "google/gemma-4-31b-it"
      },
      "models": {
        "google/gemini-3.1-pro-preview": {},
        "google/gemma-4-31b-it": {}
      },

#

      "timeoutSeconds": 18000
    }
  },

  "models": {
    "providers": {
      "google": {
        "timeoutSeconds": 600
      }
    }
  },

  "session": {
    "dmScope": "per-channel-peer"
  }

#

Or safer via CLI:

openclaw config set models.providers.google.timeoutSeconds 600
openclaw gateway restart

If it still throws the idle-timeout error, also set:

openclaw config set agents.defaults.llm.idleTimeoutSeconds 600
openclaw gateway restart

#

Basically: google is the <id>, and it goes under top-level models.providers, not inside agents.defaults.models.

slender wagon May 13, 2026, 4:32 PM

#

error:
openclaw config set models.providers.google.timeoutSeconds 600
openclaw gateway restart

🦞 OpenClaw 2026.5.7 (eeef486) — I'm like tmux: confusing at first, then suddenly you can't live without me.

Error: Config validation failed: models.providers.google.baseUrl: Invalid input: expected string, received undefined

wheat groveBOT May 13, 2026, 4:33 PM

#

Ah yep, that means config set models.providers.google.timeoutSeconds ... created a partial custom google provider, and custom providers need a baseUrl.

Set the whole provider block instead:

openclaw config set models.providers.google '{"baseUrl":"https://generativelanguage.googleapis.com/v1beta","api":"google-generative-ai","timeoutSeconds":600}' --strict-json --merge
openclaw gateway restart

That keeps google as the provider id and tells OpenClaw it’s the direct Gemini/AI Studio API, not an OpenAI-compatible endpoint.

If your config is now stuck with the partial invalid entry, the same command above should repair it by adding the missing baseUrl. If it still complains, paste the models section from ~/.openclaw/openclaw.json and we can patch the exact JSON.

Also still worth setting the idle watchdog, since your original error was “no response before idle timeout”:

#

openclaw config set agents.defaults.llm.idleTimeoutSeconds 600
openclaw gateway restart

slender wagon May 13, 2026, 4:35 PM

#

openclaw config set models.providers.google '{"baseUrl":"https://generativelanguage.googleapis.com/v1beta","api":"google-generative-ai","timeoutSeconds":600}' --strict-json --merge
openclaw gateway restart

🦞 OpenClaw 2026.5.7 (eeef486) — I'll refactor your busywork like it owes me money.

Error: Config validation failed: models.providers.google.models: Invalid input: expected array, received undefined

🦞 OpenClaw 2026.5.7 (eeef486) — I'm not magic—I'm just extremely persistent with retries and coping strategies.

#timeoutSeconds