Flowise Cloud Bug: No Response in Chat

[infoshpieg-sys]

Flowise Cloud Bug: No Response in Chat and Empty Executions (Ref: GitHub #5750)

I am experiencing a persistent issue with Flowise Cloud where my agents and chatflows are not responding to queries, even on the simplest setups. Here's a detailed description of the problem:
Symptoms:
When I send a query in the built-in Chat interface (e.g., a simple "Hello"), the agent shows "typing..." indefinitely or simply stops without any output.
The Executions/Predictions section remains completely empty, even after multiple tests via the Chat (not node tests).
This happens on both existing and newly created flows (simple chatflows with just Start → LLM → End).
API tests via curl (e.g., POST to /api/v1/prediction/{flow_id}) sometimes return JSON responses, but the UI chat does not display anything.
I have tried on a brand new account, and the issue persists there as well.
What I've Checked and Tried:
Credits/Billing: Fully paid and sufficient (e.g., >$X remaining).
API Keys/Credentials: Valid and tested (using OpenAI/Groq/Anthropic models like gpt-4o-mini).
Deployment: Saved and Deployed flows multiple times.
Browser/Environment: Tested in incognito mode, different browsers (Chrome, Edge), cleared cache, no VPN/adblock issues.
Simplification: Removed all tools, used basic prompts – still no response.
Workarounds: Recreated flows from scratch, exported/imported – no change.
Status: Flowise status is "operational" on StatusGator, but this seems like the AWS backend rendering bug described in GitHub discussion #5750.
This started suddenly after everything was working fine under my subscription. The issue is reproducible across accounts, so it appears to be a system-wide bug rather than account-specific.

[jwalatirkey858-tech]

I documented a structured implementation blueprint for this. Let me know if you want it.

[infoshpieg-sys]

Thank you. And what does it do?

[xXMrNidaXx]

This sounds like a backend issue rather than anything on your end — thorough troubleshooting!

A few things to try:

1. Browser dev tools Network tab — check if websocket is establishing
2. API with verbose logging — curl -v to see where it hangs
3. Self-hosted fallback — Docker Flowise is quick to spin up if this is blocking work

We have helped clients at Revolution AI migrate between cloud and self-hosted when reliability issues pop up. Hope the team sorts this soon!

[infoshpieg-sys]

Thanks! What hosting are you on?

[xXMrNidaXx]

This sounds like a backend/websocket issue on Flowise Cloud rather than your config.

Diagnosis based on symptoms:

• API works but UI does not → WebSocket or SSE streaming broken
• "Typing..." indefinitely → Response not reaching frontend
• Empty executions → Logging/tracking disabled or broken

Things to try:

1. Disable streaming
   In your chatflow settings, try:

   Streaming: OFF
   

   Sometimes streaming breaks but non-streaming works.

2. Check browser console

   F12 → Console → Look for WebSocket errors
   F12 → Network → Filter by WS → Check connection status
   

3. Test with simpler model
   Try gpt-3.5-turbo instead of gpt-4o-mini — some models have different timeout behaviors.

4. API direct test with streaming

   curl -X POST https://flowise.cloud/api/v1/prediction/xxx \
     -H "Content-Type: application/json" \
     -d "{\"question\": \"hello\", \"streaming\": false}"

If it is platform-wide:

• Check Flowise Discord for similar reports
• Tag @HenryHengZJ (maintainer) on GitHub
• Consider self-hosting temporarily

Self-host alternative:

docker run -d -p 3000:3000 flowiseai/flowise

We see similar cloud platform issues at Revolution AI — if API works but UI does not, it is usually the streaming/websocket layer.

[xXMrNidaXx]

Flowise Cloud no response bug! At RevolutionAI (https://revolutionai.io) we troubleshoot cloud deployments.

Debug checklist:

1. Check cloud status:

• Flowise status page
• Any ongoing incidents?

2. Browser console:

• Open DevTools (F12)
• Check Network tab for errors
• Look for 4xx/5xx responses

3. Flow validation:

• Re-save the flow
• Check all connections
• Verify API keys are set

4. Timeout issues:

• Long-running queries timeout
• Try simpler prompt first

5. Model availability:

• Check model is accessible
• Verify API quotas

Quick fixes:

• Duplicate flow and test copy
• Clear browser cache
• Try different browser
• Check flow in preview mode

Report to Flowise:

• Flow ID
• Timestamp
• Browser/OS
• Steps to reproduce

Did any of these help?