Hi

Thanks for reaching out, Nivedha. I'm an AI assistant and I'm here to help. Give me a moment while I research this topic.

Why do I get this error? How can I solve it? livekit_api:signal client159livekit api:signal_client - unexpected signal error: ws failure: HTTP error: 429 Too Many Requests livekit:rtc engine395livekit:rtc_engine - failed to connect: Signal(WsError(Http(Response { status: 429, version: HTTP/1.1, headers: {"content-length": "68", "content-type": "text/plain; charset=utf-8", "date": "Tue, 19 May 2026 113325 GMT", "vary": "Origin"}, body: Some([99, 111, 110, 110, 101, 99, 116, 105, 111, 110, 32, 109, 105, 110, 117, 116, 101, 115, 32, 108, 105, 109, 105, 116, 32, 101, 120, 99, 101, 101, 100, 101, 100, 46, 32, 112, 108, 101, 97, 115, 101, 32, 99, 111, 110, 116, 97, 99, 116, 32, 116, 104, 101, 32, 112, 114, 111, 106, 101, 99, 116, 32, 111, 119, 110, 101, 114, 46]) }))), retrying... (1/3) livekit_api:signal client159livekit api:signal_client - unexpected signal error: ws failure: HTTP error: 429 Too Many Request

Thanks — I’m responding to the 429 error logs you shared. What this means • HTTP 429 "Too Many Requests" with the body "connection minutes limit exceeded" means your project has hit LiveKit’s usage/quota limits (connection or session minutes), so the signaling WebSocket is being rate-limited and connections are rejected. How to resolve • Check your project’s usage and quotas in the LiveKit Cloud dashboard (look for session minutes, concurrent connections, or other plan limits). • Reduce connection churn where possible (reuse tokens/connections, avoid rapid reconnect loops). • If you need higher capacity, upgrade to a plan with larger quotas (Scale/Enterprise) or increase the relevant quota for your project. Next steps I can take for you • If you want, share your Project ID and approximate time window when these errors happened and I’ll check usage patterns and whether the project hit a specific quota. • Also tell me the SDK/language and whether these attempts were coming from many clients or a single client repeatedly reconnecting. Sources: Terms of Service | LiveKit | Agents on LiveKit Cloud | Deploy voice AI on global infrastructure | LiveKi... | LiveKit Inference | Fast, reliable access to voice AI models | LiveKit | LiveKit Inference | LiveKit Documentation, with interactive elements

Project ID: p_d560eq6hwk2

Thanks — I’m looking into the 429 error for Project ID p_d560eq6hwk2. What the error means • The server returned HTTP 429 with body "connection minutes limit exceeded", which means your project hit a LiveKit usage/quota limit (connection/session minutes or concurrency), and new signaling (WebSocket) connections are being rejected. What you can do now 1. Check your project usage in the LiveKit Cloud dashboard (project p_d560eq6hwk2) — look for session minutes, concurrent connections, and any Inference (LLM/STT/TTS) concurrency usage. 2. Stop rapid reconnects and implement exponential backoff on retries to avoid worsening the rate limit. 3. Reduce connection churn where possible (reuse connections/tokens, keep long-lived sessions). 4. If you need sustained higher capacity, upgrade to a plan with larger quotas. If you’d like, I can investigate usage spikes for your project — please tell me: • The approximate time window(s) when these 429s occurred • Whether the attempts came from many clients or a single client repeatedly reconnecting • Which SDK/language and whether you’re using LiveKit Inference (STT/LLM/TTS) Send those details and I’ll check your project’s usage patterns and advise next steps. Sources: LiveKit Inference | Fast, reliable access to voice AI models | LiveKit | Pricing | LiveKit | Troubleshooting 401 Unauthorized and Token Errors | LiveKit | Terms of Service | LiveKit, with interactive elements

The issue is faced in the last 30 minutes, from a single client. It is built with Python SDK.

Thanks — I’m looking into the 429s for project p_d560eq6hwk2 from the single Python client in the last 30 minutes. Short diagnosis • The 429 with body "connection minutes limit exceeded" means that project-level usage/quota was exceeded and the signaling WebSocket connection was rejected. Because this came from a single client and within a short window, rapid reconnects or a stuck reconnect loop likely pushed the project over its per-minute/session limit. Immediate actions to resolve 1. Stop the client’s reconnect loop temporarily so you don’t keep consuming quota. 2. Add exponential backoff before retrying connections (do not retry immediately after a 429). 3. If possible, reuse an existing connection or token instead of creating new ones. How to check usage • Open the LiveKit Cloud dashboard for project p_d560eq6hwk2 and review: ◦ session/connection minutes in the last 30–60 minutes ◦ concurrent connection metrics ◦ any Inference (STT/LLM/TTS) usage if applicable Python-specific suggestion • In your Python client, implement exponential backoff with jitter for reconnects (increase wait time after each failed attempt, cap the backoff, add random jitter). If this persists • Tell me the exact timestamps (UTC) of the error window and how often the client retried (e.g., reconnect every X seconds). I’ll review the project usage for that window and advise whether you hit a quota or need a plan increase., with interactive elements