compress
Caveman
Rule-based message compression, language packs, analytics and output mode controls.
compress
Caveman Engine
Rule-based message compression, language packs, analytics and output mode controls.
Requests
0
Tokens saved
0
Savings %
0%
Avg latency
0ms
Language Packs
Enable compression rules for specific languages.
info
Token Saver master switch is OFF — these settings will not affect requests until you turn it on from the Endpoint page or change it here.
Input compression
Rewrite chat history with shorter wording. Reduces input tokens by ~50%.
Compression Analytics
No compression analytics yet.
Output Mode
Control how compressed output appears
[OmniRoute Caveman Output Mode] Respond concise. Preserve technical terms, code, errors, URLs, and identifiers.
Bypass conditions: security, irreversible, clarification, order-sensitive