99.9% Service Level Agreement (SLA)
API100 commits to maintaining a Monthly Uptime Percentage of at least 99.9% for all production API endpoints, backed by automated multi-provider failover and financial service credits.
< 43 minutes allowable monthly downtime
Actual production availability
Proxy routing overhead target
1. Service Availability Calculation
Monthly Uptime Percentage is calculated as:Monthly Uptime % = ((Total Minutes in Month - Downtime Minutes) / Total Minutes in Month) * 100Downtime is defined as any continuous 5-minute interval where the API100 Gateway returns HTTP 5xx responses or connection timeouts on valid, authenticated chat completion requests.
2. Redundancy & Upstream Circuit Breaking
To insulate client applications from individual model lab outages (such as unexpected OpenAI or Anthropic degradation), API100 deploys automated circuit breakers. When an upstream model’s 5xx error rate exceeds 5% over a 30-second window, the gateway automatically redirects traffic to redundant fallback providers or cached replica endpoints.
3. Service Credit Remedies
If API100 fails to meet the 99.9% uptime commitment in any calendar billing cycle, customers are eligible for service credits applied directly to their account wallet:
- Uptime 99.0% - 99.89%: 10% credit of monthly consumption
- Uptime 95.0% - 98.99%: 25% credit of monthly consumption
- Uptime < 95.0%: 50% credit of monthly consumption