Advanced Features

Auto Failover

When your primary model provider experiences downtime, TokSpan automatically routes traffic to a backup provider serving the same model — with zero dropped requests and zero manual intervention.

How It Works

TokSpan monitors upstream provider health in real-time. When a provider returns repeated 5xx errors or exceeds a timeout threshold, traffic is automatically redirected to the next provider in your failover chain. Once the primary recovers, traffic resumes normal routing.

The failover chain is configured automatically by TokSpan across our upstream providers. No setup is required on your side — you just keep calling the same model name.

Zero configuration required. Failover is enabled automatically on every model — there is nothing to set up in your dashboard, and your code stays exactly the same.

How TokSpan Builds Your Failover Chain

For every model, TokSpan maintains a failover chain across multiple providers serving the same model. If the primary provider fails, traffic flows to the next healthy provider for that same model — no action needed from you.

PriorityModelProviderHow It's Used
1 (Primary)GPT seriesOpenAIPrimary choice — best overall quality
2 (Fallback)GPT seriesother model providersBackup route for the same model
Same-model routing: When the primary route fails, TokSpan switches to a backup route for the same model (e.g., another provider serving the GPT series). Failover never crosses to a different model — your output stays consistent while provider outages are still handled automatically.

Monitoring Failover Events

You can track failover activity through:

  • Dashboard logs — View your request history to confirm traffic switched to the backup provider
  • Webhooks — Set up real-time notifications in your account (see Webhooks)
  • Usage analytics — Per-model usage spikes indicate failover activity