Bug type
Regression (worked before, now fails)
Beta release blocker
No
Summary
Issue with webchat where Codex route returns HTML in outputs[0].text causing webchat and TUI requests to hang and queue.
Steps to reproduce
run openclaw infer model run --local --model "openai-codex/gpt-5.4" --prompt "pong" --json
Then run a control test openclaw infer model run --local --model "ollama/gemma4:31b-cloud" --prompt "pong" --json
Expected behavior
Codex route returns normal model text JSON output.
Actual behavior
Codex route returns HTML in outputs[0].text (starts with ...).
openclaw logs --follow shows repeated:
Long error truncated: <html> ...
Webchat/TUI requests hang and queue.
OpenClaw version
2026.4.14
Operating system
macOS 15.7.5
Install method
npm global
Model
openai-codex/gpt-5.4
Provider / routing chain
openclaw > openai (sorry, never sure what to add here)
Additional provider/model setup details
No response
Logs, screenshots, and evidence
`openclaw infer model run --local --model "ollama/gemma4:31b-cloud" --prompt "pong" --json \
| jq '{provider,model,outputs}'`
{
"provider": "ollama",
"model": "gemma4:31b-cloud",
"outputs": [
{
"text": "ping",
"mediaUrl": null
}
]
}
`openclaw infer model run --local --model "openai-codex/gpt-5.4" --prompt "pong" --json \
| jq '{provider,model,outputs}'`
{
"provider": "openai-codex",
"model": "gpt-5.4",
"outputs": [
{
"text": "<html>\n <head>\n <meta name=\"viewport\" content=\"width=device-width, initial-scale=1\" />\n <style global>body{font-family:Arial,Helvetica,sans-serif}.container{align-items:center;display:flex;flex-direction:column;gap:2rem;height:100%;justify-content:center;width:100%}@keyframes enlarge-appear{0%{opacity:0;transform:scale(75%) rotate(-90deg)}to{opacity:1;transform:scale(100%) rotate(0deg)}}.logo{color:#8e8ea0}.scale-appear{animation:enlarge-appear .4s ease-out}@media (min-width:768px){.scale-appear{height:48px;width:48px}}.data:empty{display:none}.data{border-radius:5px;color:#8e8ea0;text-…",
"mediaUrl": null
}
]
}Impact and severity
No response
Additional information
- This appears Codex-route-specific, not a general runtime failure.
- Separate context-meter bug already filed (this issue is specifically HTML payload from Codex route).
- Ollama route works correctly (provider: "ollama", valid text output).
Bug type
Regression (worked before, now fails)
Beta release blocker
No
Summary
Issue with webchat where Codex route returns HTML in
outputs[0].textcausing webchat and TUI requests to hang and queue.Steps to reproduce
run
openclaw infer model run --local --model "openai-codex/gpt-5.4" --prompt "pong" --jsonThen run a control test
openclaw infer model run --local --model "ollama/gemma4:31b-cloud" --prompt "pong" --jsonExpected behavior
Codex route returns normal model text JSON output.
Actual behavior
Codex route returns
HTML in outputs[0].text(starts with ...).openclaw logs --followshows repeated:Long error truncated: <html> ...Webchat/TUI requests hang and queue.
OpenClaw version
2026.4.14
Operating system
macOS 15.7.5
Install method
npm global
Model
openai-codex/gpt-5.4
Provider / routing chain
openclaw > openai (sorry, never sure what to add here)
Additional provider/model setup details
No response
Logs, screenshots, and evidence
Impact and severity
No response
Additional information