> ## Documentation Index
> Fetch the complete documentation index at: https://docs.ninjachat.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Fallback Chains

> Try models in sequence — if one fails or scores low, the next one runs.

Set `model` to a `>` separated list of model IDs. If the first model fails or scores below your quality threshold, the next one runs automatically.

## Request

```json theme={null}
{
  "model": "claude-opus-4.6>gpt-5>gemini-3.1-pro",
  "messages": [{"role": "user", "content": "Review this contract clause..."}],
  "min_quality": 0.85
}
```

## Response

```json theme={null}
{
  "model": "gpt-5",
  "fallback": {
    "chain": ["claude-opus-4.6", "gpt-5", "gemini-3.1-pro"],
    "triggered": true,
    "attempts": [
      {"model": "claude-opus-4.6", "success": false, "error": "timeout"},
      {"model": "gpt-5", "success": true, "quality_score": 0.94}
    ]
  },
  "cost": {"this_request": "$0.006"}
}
```

## Parameters

| Parameter           | Type    | Default | Description                                                              |
| ------------------- | ------- | ------- | ------------------------------------------------------------------------ |
| `model`             | string  | —       | Up to 4 models separated by `>`. Example: `"a>b>c"`                      |
| `min_quality`       | number  | —       | If the response scores below this (0–1), fall back to the next model.    |
| `fallback_on_error` | boolean | `true`  | Continue to the next model on error. Set to `false` to fail immediately. |

## Billing

You're charged for each model that runs successfully. If the first model times out and the second succeeds, you pay only for the second.

## Code examples

<CodeGroup>
  ```python Python theme={null}
  import requests, os

  r = requests.post("https://www.ninjachat.ai/api/v1/chat",
      headers={"Authorization": f"Bearer {os.environ['NINJACHAT_API_KEY']}"},
      json={
          "model": "claude-opus-4.6>gpt-5>gemini-3.1-pro",
          "messages": [{"role": "user", "content": "Analyze this legal document..."}],
          "min_quality": 0.85,
      }
  )
  data = r.json()
  print(f"Resolved to: {data['model']}")
  print(f"Fallback triggered: {data['fallback']['triggered']}")
  ```

  ```bash cURL theme={null}
  curl -X POST https://www.ninjachat.ai/api/v1/chat \
    -H "Authorization: Bearer nj_sk_YOUR_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
      "model": "claude-opus-4.6>gpt-5>gemini-3.1-pro",
      "messages": [{"role": "user", "content": "Summarize this research paper..."}],
      "min_quality": 0.8
    }'
  ```
</CodeGroup>