Workflows and features
Publishing changes safely, provider fallbacks, structured JSON, conversation memory and caching.
5 articles
- Does saving a workflow change what production uses?No. Saving changes only the Draft; production runs the Live version until you publish. Learn how ModelRiver Draft, Live, publish and rollback work.
- What happens to my app when an AI provider goes down?Add backup models to a ModelRiver workflow and requests fail over automatically when a provider errors, times out or rate-limits. Here's how failover works.
- How do I make sure the AI always returns valid JSON?Attach a structured output (JSON schema) to a ModelRiver workflow and every response is checked against it, with backups tried if it fails. Here's how.
- Do I have to send the whole chat history with every message?No. Turn on session memory in a ModelRiver workflow and send only the new message with a session_id. Learn how conversation memory works and how to reset it.
- Does response caching save me money?Caching in ModelRiver returns identical repeat requests in under 10ms without calling the provider, saving tokens. Cache hits still count as requests.