Issue Station
Sign In
Tools hereToolbox

openclaw - 💡(How to fix) Fix [Bug]: GatewayClientRequestError: FailoverError: LLM request timed out.

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…

OpenClaw, using local LM Studio, gives timeout very fast. In around 3 seconds. I've increased all timeout parameters, but reading the documentation, those should be by default 30 or more seconds.

Root Cause

OpenClaw, using local LM Studio, gives timeout very fast. In around 3 seconds. I've increased all timeout parameters, but reading the documentation, those should be by default 30 or more seconds.

Code Example

│
◇

OpenClaw 2026.5.26 (10ad3aa) — No $999 stand required.

model.run via local
provider: lmstudio
model: qwen3.5-4b
outputs: 1
Hello! It looks like you're testing something here. How can I help you

real    0m6.797s
user    0m2.628s
sys     0m0.369s

---

timestamp: 5/28/2026, 2:11:41 PM
type: llm.prediction.output
modelIdentifier: qwen3.5-4b
output:
<think>
The user is asking me to continue the conversation from the previous turn. However, I notice that there's no actual question or request in this message - it just says "test". This appears to be a test message or perhaps an incomplete input.

Since this seems like a test message without any specific content, I should respond appropriately by acknowledging the test and offering assistance.
</think>

Hello! It looks like you're testing something here. How can I help you today? Feel free

---
RAW_BUFFERClick to expand / collapse

Bug type

Behavior bug (incorrect output/state without crash)

Beta release blocker

No

Summary

OpenClaw, using local LM Studio, gives timeout very fast. In around 3 seconds. I've increased all timeout parameters, but reading the documentation, those should be by default 30 or more seconds.

Steps to reproduce

  1. $ time openclaw infer model run --local --model lmstudio/qwen3.5-4b --prompt "test"
│
◇

OpenClaw 2026.5.26 (10ad3aa) — No $999 stand required.

model.run via local
provider: lmstudio
model: qwen3.5-4b
outputs: 1
Hello! It looks like you're testing something here. How can I help you

real    0m6.797s
user    0m2.628s
sys     0m0.369s

When I check the logs from lms I can see that it wanted to answer more, but was cut short:

timestamp: 5/28/2026, 2:11:41 PM
type: llm.prediction.output
modelIdentifier: qwen3.5-4b
output:
<think>
The user is asking me to continue the conversation from the previous turn. However, I notice that there's no actual question or request in this message - it just says "test". This appears to be a test message or perhaps an incomplete input.

Since this seems like a test message without any specific content, I should respond appropriately by acknowledging the test and offering assistance.
</think>

Hello! It looks like you're testing something here. How can I help you today? Feel free

Expected behavior

OpenClaw should wait longer for the request.

Actual behavior

Local models are unusable.

OpenClaw version

2026.5.26

Operating system

Fedora 44

Install method

npm global

Model

qwen3.5-4b

Provider / routing chain

openclaw -> local lm studio

Additional provider/model setup details

No response

Logs, screenshots, and evidence

Impact and severity

No response

Additional information

No response

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

OpenClaw should wait longer for the request.

Still need to ship something?

×6

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

Back to top recommendations

TRENDING