Glossary · Inference & serving

KV cache

The stored key and value tensors from previous tokens that let a transformer avoid recomputing attention over the whole sequence at each step.

Related terms

• FlashAttention
• Ollama
• PagedAttention
• SGLang
• autoscaling (inference)
• batch API

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.