openclaw - 💡(How to fix) Fix [Bug]: Possible bug: OpenRouter probe succeeds, but no actual response is returned [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#69828Fetched 2026-04-22 07:47:51
View on GitHub
Comments
0
Participants
1
Timeline
2
Reactions
0
Participants
Timeline (top)
labeled ×2

I think I may be seeing a bug on Windows with OpenClaw 2026.4.15.

OpenRouter auth/probe seems to work, but actual responses fail.

When I run:

openclaw infer model run --model openrouter/anthropic/claude-3.7-sonnet --prompt "Say only OK" --json

or:

openclaw infer model run --gateway --model openrouter/anthropic/claude-3.7-sonnet --prompt "Say only OK" --json

I get: "⚠️ Agent couldn't generate a response. Please try again."

On local transport I also saw: "incomplete turn detected" and payloads=0

So the configuration seems valid, but OpenClaw does not return the actual model response. This looks like a bug rather than a setup mistake.

Root Cause

I think I may be seeing a bug on Windows with OpenClaw 2026.4.15.

OpenRouter auth/probe seems to work, but actual responses fail.

When I run:

openclaw infer model run --model openrouter/anthropic/claude-3.7-sonnet --prompt "Say only OK" --json

or:

openclaw infer model run --gateway --model openrouter/anthropic/claude-3.7-sonnet --prompt "Say only OK" --json

I get: "⚠️ Agent couldn't generate a response. Please try again."

On local transport I also saw: "incomplete turn detected" and payloads=0

So the configuration seems valid, but OpenClaw does not return the actual model response. This looks like a bug rather than a setup mistake.

RAW_BUFFERClick to expand / collapse

Bug type

Regression (worked before, now fails)

Beta release blocker

No

Summary

I think I may be seeing a bug on Windows with OpenClaw 2026.4.15.

OpenRouter auth/probe seems to work, but actual responses fail.

When I run:

openclaw infer model run --model openrouter/anthropic/claude-3.7-sonnet --prompt "Say only OK" --json

or:

openclaw infer model run --gateway --model openrouter/anthropic/claude-3.7-sonnet --prompt "Say only OK" --json

I get: "⚠️ Agent couldn't generate a response. Please try again."

On local transport I also saw: "incomplete turn detected" and payloads=0

So the configuration seems valid, but OpenClaw does not return the actual model response. This looks like a bug rather than a setup mistake.

Steps to reproduce

  1. Configure OpenRouter in OpenClaw 2026.4.15 on Windows.
  2. Confirm that model probe/auth succeeds for openrouter/anthropic/claude-3.7-sonnet.
  3. Run: openclaw infer model run --model openrouter/anthropic/claude-3.7-sonnet --prompt "Say only OK" --json
  4. Observe that OpenClaw does not return the real model response and instead returns: ⚠️ Agent couldn't generate a response. Please try again.
  5. Run: openclaw infer model run --gateway --model openrouter/anthropic/claude-3.7-sonnet --prompt "Say only OK" --json
  6. Observe the same fallback response again.
  7. On local transport, also observe log output showing incomplete turn detected and payloads=0.

Expected behavior

OpenClaw should return the actual model output, for example: OK

Actual behavior

OpenClaw returns the fallback message instead of the real model response: ⚠️ Agent couldn't generate a response. Please try again.

On local transport I also saw: incomplete turn detected and payloads=0

OpenClaw version

2026.4.15 (041266a)

Operating system

Windows 11

Install method

pnpm

Model

openrouter/anthropic/claude-3.7-sonnet

Provider / routing chain

openclaw -> openrouter -> anthropic/claude-3.7-sonnet

Additional provider/model setup details

OpenRouter authentication/probe succeeds, so the API key and model reference appear valid.

The problem happens when running actual inference. Both local and gateway execution return the fallback message instead of the real model output.

Tested commands:

openclaw infer model run --model openrouter/anthropic/claude-3.7-sonnet --prompt "Say only OK" --json

openclaw infer model run --gateway --model openrouter/anthropic/claude-3.7-sonnet --prompt "Say only OK" --json

In the local case I also saw: incomplete turn detected payloads=0

Logs, screenshots, and evidence

Impact and severity

No response

Additional information

No response

extent analysis

TL;DR

The issue might be related to a bug in OpenClaw 2026.4.15 on Windows, causing it to fail to return the actual model response, and a workaround could involve checking the model configuration or trying a different version of OpenClaw.

Guidance

  • Verify that the model openrouter/anthropic/claude-3.7-sonnet is correctly configured and supported by OpenClaw 2026.4.15.
  • Check the OpenClaw documentation for any known issues or limitations on Windows that might be causing the problem.
  • Try running the same command with a different model to see if the issue is specific to openrouter/anthropic/claude-3.7-sonnet.
  • Consider trying a different version of OpenClaw to see if the issue is version-specific.

Notes

The issue seems to be specific to OpenClaw 2026.4.15 on Windows, and the fact that authentication/probe succeeds suggests that the API key and model reference are valid. However, without more information or logs, it's difficult to provide a more specific solution.

Recommendation

Apply workaround: Try a different version of OpenClaw or a different model to see if the issue persists, as this might help identify if the problem is specific to the current version or model configuration.

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 return the actual model output, for example: OK

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]: Possible bug: OpenRouter probe succeeds, but no actual response is returned [1 participants]