Reference

Hober-Specific Features

Model Fallbacks

Pass an array of models in the models field. Hober tries each in order until one succeeds. If the primary model is down or rate-limited, the next model is used automatically.

{
  "models": [
    "deepseek/deepseek-chat-v3.2",
    "qwen/qwen3-235b",
    "glm/glm-4-plus"
  ],
  "messages": [{"role": "user", "content": "Hello"}]
}

Provider Preferences

Control which providers are used for routing with the provider field.

provider.sortstringoptional

"price" (cheapest first) or "latency" (fastest first)

provider.onlyarrayoptional

Only use these providers: ["deepseek", "qwen"]

provider.ignorearrayoptional

Never use these providers: ["glm"]

{
  "model": "deepseek/deepseek-chat-v3.2",
  "messages": [{"role": "user", "content": "Hello"}],
  "provider": {
    "sort": "latency",
    "ignore": ["glm"]
  }
}

Model Suffixes

Append routing hints to any model slug:

SuffixBehaviorExample
:nitroPrefer lowest-latency provider endpointdeepseek/deepseek-chat-v3.2:nitro
:floorPrefer cheapest provider endpointqwen/qwen3-235b:floor