LLM Playground
LLM Routing Playground
Compare responses, latency, and token costs across four models. Running in demo mode — set LITELLM_PROXY_URL to connect to a live model fleet.
Available models
GPT-4o Mini$0.15/$0.6
OpenAI · Fast
Fast, low-cost tasks
DeepSeek V3$0.27/$1.1
DeepSeek · Fast
Cost-sensitive production
Claude Sonnet 4$3/$15
Anthropic · Moderate
Long-context reasoning
Gemini 2.0 FlashFree
Google · Fast
Free tier, high volume
What you’re looking at
LatencyEnd-to-end, including simulated network overhead
TokensInput + output, rounded
CostPer-request at current token counts
ModeDemo = synthetic. Set LITELLM_PROXY_URL for live calls.