openclaw - 💡(How to fix) Fix TUI startup extremely slow: model-resolution 11-16s per message + event loop blocked for 5-16 minutes [2 comments, 3 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#75668Fetched 2026-05-02 05:32:00
View on GitHub
Comments
2
Participants
3
Timeline
10
Reactions
6
Author
Timeline (top)
cross-referenced ×7commented ×2subscribed ×1

Fix Action

Fix / Workaround

[trace:embedded-run] startup stages:
  workspace: 0ms
  runtime-plugins: 0-1ms
  hooks: 0ms
  model-resolution: 11010ms ← always 11-16 seconds
  auth: 534ms
  attempt-dispatch: 691ms
  total: 12236ms

Code Example

[trace:embedded-run] startup stages:
  workspace: 0ms
  runtime-plugins: 0-1ms
  hooks: 0ms
  model-resolution: 11010ms ← always 11-16 seconds
  auth: 534ms
  attempt-dispatch: 691ms
  total: 12236ms

---

liveness warning: eventLoopDelayMaxMs=999116.8  (16 minutes!)
liveness warning: eventLoopDelayMaxMs=482646.9  (8 minutes)
liveness warning: eventLoopDelayMaxMs=310311.4  (5 minutes)
RAW_BUFFERClick to expand / collapse

Hi, I'm experiencing very slow TUI startup with xiaomi/mimo-v2.5-pro. Two issues:

Issue 1: model-resolution takes 11-16s on every message

[trace:embedded-run] startup stages:
  workspace: 0ms
  runtime-plugins: 0-1ms
  hooks: 0ms
  model-resolution: 11010ms ← always 11-16 seconds
  auth: 534ms
  attempt-dispatch: 691ms
  total: 12236ms

This happens on every message, not just the first one. Network is confirmed fine — the Xiaomi API endpoint responds in ~156ms (ping 0.4ms). The model-resolution stage alone accounts for ~90% of startup time.

Issue 2: Event loop blocked for minutes

The gateway's event loop gets blocked for extremely long periods:

liveness warning: eventLoopDelayMaxMs=999116.8  (16 minutes!)
liveness warning: eventLoopDelayMaxMs=482646.9  (8 minutes)
liveness warning: eventLoopDelayMaxMs=310311.4  (5 minutes)

During these blocks, the TUI shows "not connected to gateway" and becomes completely unresponsive. Sometimes it takes 5+ minutes just to open the chat interface.

Environment

  • OpenClaw: 2026.4.29 (a448042)
  • OS: macOS 26.4 (arm64), Mac mini
  • Model: xiaomi/mimo-v2.5-pro
  • Provider URL: https://token-plan-cn.xiaomimimo.com/anthropic
  • Node: v22.22.1
  • Network: API responds in ~156ms, ping 0.4ms (not a network issue)

Questions

  1. Is the model-resolution stage doing model capability probing on every request? If so, could results be cached?
  2. Is the event loop blocking related to model-resolution, or a separate issue?
  3. Is this specific to the Xiaomi provider or a general issue?

Happy to provide more logs if needed.

extent analysis

TL;DR

Investigate caching model capability probing results to reduce the model-resolution stage time and potentially alleviate event loop blocking.

Guidance

  • Verify if the model-resolution stage is indeed performing model capability probing on every request by checking the code or documentation related to the xiaomi/mimo-v2.5-pro model.
  • Check if there are any existing caching mechanisms for model capability probing results that can be leveraged or enabled.
  • Analyze the correlation between the model-resolution stage and event loop blocking to determine if they are related issues.
  • Consider collecting additional logs or metrics to better understand the event loop blocking issue and its potential causes.

Notes

The provided information suggests that the model-resolution stage is a significant contributor to the slow TUI startup time, but the root cause of the event loop blocking issue is still unclear. Further investigation and logging may be necessary to determine the best course of action.

Recommendation

Apply workaround: Investigate and implement caching for model capability probing results, as this may help reduce the model-resolution stage time and potentially alleviate event loop blocking.

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…

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 TUI startup extremely slow: model-resolution 11-16s per message + event loop blocked for 5-16 minutes [2 comments, 3 participants]