Total Savings
Tokens Saved
Compression Quality
Overhead
TTFB s avg
Without Headroom
Input cost
vs
Input cost without Headroom
Compression saved
Prefix Cache Impact
Cache Reads
Cache Writes
Hit Rate
Cache Busts
Providers
with cache data
Cache Efficiency
Reads (discounted)
Writes
Uncached
Per-Provider Breakdown
What Headroom Removed
No waste signals detected yet. Data appears after requests are processed.
Savings Over Time
Trend data will appear after multiple requests.
Token Usage
Before Compression
After Compression (sent)
Output Tokens
Providers
No requests yet
Performance
Headroom Overhead
Overhead Range
TTFB (upstream)
TTFB Range
Failed Requests
Pipeline Breakdown
Per-Model Token Savings
Exact tokens saved per model
| Model | Requests | Tokens Saved | Tokens Sent | Reduction |
|---|---|---|---|---|
Recent Requests
Last 10 — click row to expand
| Time | Model | Input | Output | Saved | Quality | Latency | |
|---|---|---|---|---|---|---|---|
|
Original Tokens
Compressed Tokens
Tokens Removed
Optimization Time
Transforms Applied
Waste Detected
|
|||||||
| No requests yet. Start using the proxy to see activity here. | |||||||