How-To Series · Episode 53 / 59 · Module 8: Provider Plumbing

Hermes · Fallback Providers

Primary down? Hermes switches to a backup automatically, mid-chat, without dropping your conversation.

After this videoYou can now survive provider outages without lifting a finger.

A fallback is a backup provider:model. When your main one starts failing, Hermes switches to it automatically, mid-session, and the switch is invisible: history, tool calls, and context all carry over, so the agent continues where it left off. It triggers on the failures that matter, rate limits (429) and server errors (5xx) after retries, and auth (401/403) or not-found (404) errors immediately. It's per-turn: every new message tries your primary first, so one blip doesn't pin you to the backup. Set it up with hermes fallback (interactive add/list/remove, reuses the model picker) or a fallback_providers: list in config. Covers your CLI and messaging gateway; subagents and cron run on a fixed provider.

About these resources. Every command comes from the Fallback Providers doc.

Sources · What this video distills

1 docs page · every command below traces to one of them
Primary · triggers, hermes fallback, fallback_providers config, per-turn behavior, scope
Fallback Providers
Read ↗

Commands shown · Copy and paste

each shows the source doc it came from
Interactive managerfrom source ↗
hermes fallback # add / list / remove / clear
Config (config.yaml)from source ↗
fallback_providers: [ { provider: openrouter, model: anthropic/claude-sonnet-4 } ]

Going deeper · Related Hermes docs

further reading · not sources of facts shown above

Next in the series · Episodes that build on this

E54
Credential Pools
E52
Provider Routing
E50
The Provider Landscape