Glossary · Inference & serving

prompt caching

Reusing the computed KV cache for a shared prompt prefix across multiple requests to skip redundant computation and cut cost and latency.

Related terms

• FlashAttention
• KV cache
• Ollama
• PagedAttention
• SGLang
• autoscaling (inference)

Vynaris is an inference gateway that routes each request to the cheapest right-sized model and shows the receipt. Get an API key or read the docs.