Adaptive automatically switches to backup providers when your primary provider fails. No downtime, no interruptions.

​
Fallback Configuration

Fallback is enabled by default with parallel mode for maximum reliability. You can customize or disable it as needed.

​
Default Behavior (Enabled)

const completion = await openai.chat.completions.create({
  model: "",
  messages: [{ role: "user", content: "Hello" }]
  // No fallback configuration = enabled with parallel mode
});

​
Customize Fallback

const completion = await openai.chat.completions.create({
  model: "",
  messages: [{ role: "user", content: "Hello" }],
  fallback: {
    enabled: true,
    mode: "sequential"
  }
});

​
Disable Fallback

const completion = await openai.chat.completions.create({
  model: "",
  messages: [{ role: "user", content: "Hello" }],
  fallback: {
    enabled: false
  }
});

​
Fallback Modes

​
Parallel (Default)

Try multiple providers at once for maximum speed and reliability: 
const completion = await openai.chat.completions.create({
  model: "",
  messages: [{ role: "user", content: "Hello" }],
  fallback: {
    enabled: true,
    mode: "parallel"
  }
});
Pros: Ultra-low latency, max reliability
Cons: Higher cost (multiple calls)

​
Sequential

Try providers one by one: 
const completion = await openai.chat.completions.create({
  model: "",
  messages: [{ role: "user", content: "Hello" }],
  fallback: {
    enabled: true,
    mode: "sequential"
  }
});
Pros: Lower cost, predictable
Cons: Higher latency on failures

​
Performance

  • 99.95% uptime across providers
  • Under 500ms failover latency
  • Under 30s recovery time

​
Error Handling

try {
  const completion = await openai.chat.completions.create({
    model: "",
    messages: [{ role: "user", content: "Hello" }],
    fallback: {
      enabled: true,
      mode: "sequential"
    }
  });
} catch (error) {
  if (error.code === 'all_providers_failed') {
    console.log('All providers unavailable');
  }
}