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

# Models and Pricing

> Supported model IDs and endpoint pricing for LLMLayer.

## Supported Answer Models

| Model | Pricing style | Best for |
| - | - | - |
| `llmlayer-web` | Flat `$0.007 × max_queries` | Default web-grounded answers |
| `llmlayer-fast` | Flat `$0.009 × max_queries` | Fast answers |
| `openai/gpt-4o-mini` | Token priced + LLMLayer fee | Budget token-priced answers |
| `openai/gpt-5.1` | Token priced + LLMLayer fee | Premium reasoning |

<Note>
  The `openai/gpt-5.1` backend mapping is intentionally left for your later backend update.
</Note>

## Provider Keys

`provider_key` is accepted for backward compatibility but currently ignored; provider usage is not billed to the user's provider account.

Use LLMLayer response fields and the dashboard ledger as the billing source of truth.

## Answer Pricing

### LLMLayer flat-price models

```text theme={null}
llmlayer-web  = $0.007 × max_queries
llmlayer-fast = $0.009 × max_queries
```

### Token-priced models

```text theme={null}
total = ($0.004 × max_queries)
      + model input token cost
      + model output token cost
      + optional image search cost
```

Current token prices:

| Model | Input | Output |
| - | -: | -: |
| `openai/gpt-4o-mini` | `$0.15 / 1M tokens` | `$0.60 / 1M tokens` |
| `openai/gpt-5.1` | `$1.25 / 1M tokens` | `$10.00 / 1M tokens` |

## Utility Endpoint Pricing

| Endpoint | Cost |
| - | -: |
| Web Search | `$0.002` per request |
| Scraper | `$0.001` per supported format |
| Extract `json`, `summary`, `qa` | `$0.005` per selected AI mode |
| Extract `links` | `$0.001` |
| Extract `brand` | `$0.002` |
| Crawl | `$0.001` per successful page, reported in the usage event |
| Map | `$0.002` per request |
| YouTube Transcript | `$0.003` per request |
| PDF Content | `$0.002` per request |

## Cost Control

* Start Answer requests with `llmlayer-web` and `max_queries=1`.
* Use Web Search + Scraper when you need to select sources before asking an LLM.
* Use Extract only for fields you actually need.
* Request one scrape format unless you need multiple outputs.
* Read `cost`, `llmlayer_cost`, and `model_cost` from responses when present.

## FAQ

<AccordionGroup>
  <Accordion title="Can I bring my own provider API key?">
    `provider_key` is accepted for backward compatibility but currently ignored; provider usage is not billed to the user's provider account.
  </Accordion>

  <Accordion title="What happens if I pass an unsupported model?">
    The Answer backend currently falls back to `llmlayer-web` rather than failing the request.
  </Accordion>

  <Accordion title="Where should I verify final charges?">
    Use the response cost fields for request-level visibility and the dashboard ledger as the account-level billing source of truth.
  </Accordion>
</AccordionGroup>

## Next Steps

<CardGroup cols={2}>
  <Card title="Answer API" icon="sparkles" href="/answer">
    Model selection, streaming, and response fields.
  </Card>

  <Card title="Errors & Refunds" icon="triangle-exclamation" href="/errors">
    Error shapes, status codes, and refund behavior.
  </Card>
</CardGroup>


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.