Token Usage Visualizer
Analyze your LLM logs for cost spikes and usage patterns.
Upload Usage Logs
Drag & drop your JSON/CSV logs here, or click to browse
Supported formats: .json, .csv, .log
Analyzing:
production_logs_v2.json
Total Tokens
2.4M
12%
vs yesterday
Est. Cost
$42.80
Avg $1.78/hour
Peak Usage
14:00
320k tokens/hr
Requests
8,450
~352 per hour
Token Usage Over Time (Hourly)
Cost by Model
Automated Insights
Cost Spike Detected
Unusually high usage on GPT-4 at 14:00 (+$12.50). Verify if this was expected traffic.
Optimization Opportunity
35% of requests use GPT-4 but have low complexity. Moving to GPT-4o mini could save ~$15/day.
Stable Performance
Error rates remain below 0.1%. Latency is stable at ~450ms avg.