Interactive AI Model Testing & Chat Arena
Live Multi-Model RouterInstantly choose which AI model executes queries. Kawa preserves 100% of memory and venue facts across all models.
psychology
Kawa Persistent Memory Engine
Active
Qdrant Vector Recall • MongoDB Rules • Venue Turn Histories
Live Estate Engine
Qwen 3.8 (27B Cloud)
tune
Target AI Model:
Hetzner Cloud • 262k context window
Target Model: Hetzner AI — Qwen 3.8 (27B Cloud)
Hetzner Cloud
verified_user
Kawa AI Model Playground & Operations Copilot
You are connected to Kawa's live inference router. Choose any AI model above, type your question or test prompt below, and inspect the real output, response latency, and reasoning chain.
bolt Presets:
Press Enter to send • Shift + Enter for newline
0 chars
compare Side-by-Side Multi-Model Arena
Broadcast the exact same prompt simultaneously to multiple AI engines to compare response speed, accuracy, and reasoning tokens.
Compare:
Select models above, enter a question, and click "Run Arena Compare" to see side-by-side responses.
monitor_heart AI Engine Health & Ping Diagnostic Matrix
Last tested: Click 'Run Ping Diagnostic' to refresh
Loading AI engine diagnostics...