Configuration

Out of the box, no configuration is needed — pricing comes from LiteLLM automatically. Publish the config only if you want to override prices or tune the cache:

php artisan vendor:publish --tag=ai-costs-config

This creates config/ai-costs.php:

return [

    'litellm' => [
        'url' => env(
            'AI_COSTS_LITELLM_URL',
            'https://raw.githubusercontent.com/BerriAI/litellm/main/model_prices_and_context_window.json',
        ),
        'cache_ttl' => env('AI_COSTS_CACHE_TTL', 86400),
    ],

    'models' => [
        // Local overrides — these take precedence over LiteLLM.
        // Format: 'model-name' => ['input' => price_per_1M, 'output' => price_per_1M],
    ],

];

Local pricing overrides

Add entries under models to pin a price or define a custom model that LiteLLM doesn't know about:

'models' => [
    'my-custom-model' => ['input' => 1.00, 'output' => 3.00],
    'gpt-4o'          => ['input' => 2.50, 'output' => 10.00],
],

Prices are USD per 1M tokens, matching the LiteLLM convention.

Key normalisation

Model names have dots removed for config-key compatibility:

Model	Config key
gpt-4.1	gpt-41
claude-sonnet-4-6	claude-sonnet-4-6
gemini-1.5-pro	gemini-15-pro

The calculator handles this normalisation automatically — you don't need to think about it at the call site.

Wildcard / prefix matching

A trailing * makes the entry match any model whose normalised key starts with the prefix:

'models' => [
    'my-custom*' => ['input' => 1.00, 'output' => 3.00],
],

This matches my-custom-model, my-custom-v2, and so on.

LiteLLM cache

LiteLLM pricing is cached for 24 hours by default. Tune via env:

AI_COSTS_CACHE_TTL=86400

Or change the source URL (e.g. to a fork or mirror):

AI_COSTS_LITELLM_URL=https://example.com/my-mirror.json

To clear the cache manually (e.g. after pushing a fix to the LiteLLM repo):

use Aaix\LaravelAiCosts\Support\LitellmPricingProvider;

LitellmPricingProvider::clearCache();