#Update issue
1 messages · Page 1 of 1 (latest)
The important line is this one:
SsrFBlockedError ... Blocked: resolves to private/internal/special-use IP address for http://beelink:1234.
So 2026.5.12 is reaching your LM Studio/OpenAI-compatible provider through the guarded model fetch path, and the SSRF protection is blocking the LAN/private hostname. Since you trust that local model host, add the explicit provider opt-in:
openclaw config set models.providers.lmstudio.request.allowPrivateNetwork true --strict-json
openclaw gateway restart
openclaw models status
If your provider id is not actually lmstudio, use that id instead: models.providers.<id>.request.allowPrivateNetwork.
Docs: https://docs.openclaw.ai/providers/lmstudio and https://docs.openclaw.ai/gateway/local-models
Two side notes from the log: