Total Requests
2,842,910
Successful Requests
2,812,650
Failed Requests
30,260
Average Latency
238ms

Runtime Overview

Real-time gateway concurrency, throughput rate, and execution telemetry.
Live Syncing
Concurrent Requests
342
Active WebSocket & HTTP streams
Tokens Per Minute
485.2K
324K in / 161.2K out
Success Rate (SLA)
99.12%
SLA Target > 99.00%
Average Response Time
215ms
Cache Hit: 42ms

Requests Per Minute

24-hour continuous inference traffic throughput
Peak: 4,420 req/m

Quality & Cost Metrics

Realtime
Quality Score 98.4%
Hallucination Rate: 0.8% Safety Guard: 100%
Spend MTD $18,420
Budget: $25,000 73.6% Consumed
Verified Business Savings +$64,200

Latency Distribution

P95 latency vs P50 median across all model providers

Capability Usage

Task breakdown by generation type

AI Issues & Insights Detected

High Priority

P95 latency spike detected in Customer Care Co-Pilot. Prompt token size increased by 42% due to uncompressed RAG chunks.

View All Insights

Applications Overview

High-throughput active applications running on the Orvexa platform
Manage All Applications
Application Version Environment Throughput P95 Latency Quality Score Status Action
Customer Care Co-Pilot
claude-3-5-sonnet
v2.4.1 Production 1,420 req/m 280ms 98.4% Healthy Details
Legal Doc Synthesizer
gpt-4o
v1.9.0 Production 680 req/m 210ms 99.1% Healthy Details
FraudGuard Vision Engine
gemini-1.5-pro
v3.0.2 Production 920 req/m 185ms 99.8% Healthy Details
CodeAssist Enterprise
claude-3-5-sonnet
v2.1.0 Canary 410 req/m 315ms 97.9% Investigating Details

Live Requests Stream

Open Realtime Inspector
Request ID Time Application Model Input Tokens Output Tokens Latency Status Trace
REQ-98421 01:52:14 Customer Care Co-Pilot claude-3-5-sonnet 1,240 420 218ms 200 OK Trace
REQ-98420 01:52:12 FraudGuard Vision gemini-1.5-flash 850 120 92ms 200 OK Trace
REQ-98419 01:52:09 Legal Doc Synthesizer gpt-4o 3,820 980 380ms 200 OK Trace
REQ-98418 01:52:05 CodeAssist Enterprise claude-3-5-sonnet 4,120 0 820ms 429 Rate Limit Trace