Routing — ML Junction docs

Routing begins by eliminating inactive, disabled, unhealthy, capability-incompatible, privacy-incompatible, and policy-incompatible routes. The strategy then orders the remaining candidates. Fallback policy controls what may happen after a failed attempt.

Strategy · Behavior
balanced · Balances route quality, health, cost, and latency
quality · Prioritizes stronger and more reliable routes
cost · Prioritizes the lowest active verified price
latency · Prioritizes recent route latency
pinned · Requires routing.route or routing.provider

Routing controls

Field · Meaning
route · Pin or prefer an exact route ID
provider · Pin or prefer a provider
only_providers · Allow-list of providers
ignored_providers · Deny-list of providers
region · Requested provider region
strict_params · Disqualify routes that cannot honor requested parameters
require_zdr · Require a route where zero-data-retention can be enforced
allow_provider_training · Permit routes whose policy allows provider training
max_platform_charge_usd · Maximum platform-funded charge
max_byok_fee_usd · Maximum platform routing fee for BYOK

{
  "model": "gpt-5.5",
  "messages": [
    { "role": "user", "content": "Generate the release decision." }
  ],
  "routing": {
    "strategy": "pinned",
    "production_mode": true,
    "route": "openai-gpt-5.5-global-standard",
    "mode": "platform_only",
    "max_platform_charge_usd": 0.25,
    "strict_params": true,
    "fallbacks": {
      "provider": true,
      "model": false,
      "max_total_attempts": 2,
      "max_retries_per_route": 1,
      "preserve_context_window": true,
      "preserve_privacy": true
    }
  }
}

production_mode requires an explicit provider, route, or provider allow-list. It also requires relevant charge caps, enables strict parameter handling, and preserves context and privacy across fallback.

Canonical URL: https://mljunction.com/docs/routing