LLM Streaming

Definitions

1. The incremental delivery of large language model output as it is generated rather than only after completion finishes. LLM streaming improves perceived responsiveness but can complicate moderation, formatting, and downstream consumption.

   In plain English: Receiving large language model output bit by bit as it is generated.

   Example: "The chat UI used LLM streaming so analysts could begin reading the answer while the rest of the explanation was still arriving."

Related Terms

• AI Streaming
• LLM Benchmark
• LLM Cache
• LLM Calling
• LLM Completion
• LLM Config
• LLM Memory
• LLM Optimization
• Model Streaming