ModelRefs / Multi-Turn Conversation — AI Glossary
Multi-Turn Conversation — AI Glossary
A dialogue spanning multiple user-assistant exchanges, requiring the model to maintain coherence, context, and task state across turns.
Overview
Multi-turn evaluation tests coherence, instruction memory (following earlier constraints in later turns), and graceful handling of topic shifts. Challenging for models due to context length limits and positional biases. ChatML (OpenAI), Anthropic Human/Assistant, and Google Gemini formats define the standard message-role conventions.
Reference details
| Topic | prompting |
|---|---|
| Last reviewed | 2026-06-24 |
Related terms
Commonly confused with
Describes the interaction, not the payload. Conversation history is the mechanism that makes it possible, since chat endpoints keep no state between calls. Evaluating a model on single-turn benchmarks says little here: the failures specific to multi-turn are forgetting a constraint set several turns earlier and mishandling a topic shift, and neither is visible in a one-shot test.
Continue your research
Use these connected ModelRefs sections to compare alternatives, inspect implementation paths, and review the evidence and governance boundaries relevant to Multi-Turn Conversation — AI Glossary.
Frequently asked questions
What is Multi-Turn Conversation?
A dialogue spanning multiple user-assistant exchanges, requiring the model to maintain coherence, context, and task state across turns.
What concepts are related to Multi-Turn Conversation?
Closely related concepts include conversation history, message threading, context management.