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.
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.
| Priority | Model | Provider | How It's Used |
|---|---|---|---|
| 1 (Primary) | GPT series | OpenAI | Primary choice — best overall quality |
| 2 (Fallback) | GPT series | other model providers | Backup route for the same model |
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