Prompt Caching Inoziva Provider

Superdav AI Agent v1.12.0 inounza provider-aware prompt caching, iyo inogadzirisa mutengo weAPI nekumhanya nekuchengetedza maprompt pakati peLLM providers vakasiyana. Provider mumwe nemumwe ane nzira nemagadzirirwo akasiyana e caching.

Pfupiso

Prompt caching inokubvumira kuti:

• Uchengetedze maprompt makuru, anoshandiswa kazhinji
• Uderedze mutengo weAPI nekudzivisa kugadzirisa kwakapetwa
• Uvandudze kumhanya kwezvikumbiro zvakachengetedzwa mu cache
• Utungamirire cache lifecycle zvakajeka

Providers vakasiyana vanoita caching nenzira dzakasiyana:

• Google Gemini: cachedContents API
• Azure OpenAI: Prompt caching ine TTL
• OpenRouter: Caching yakanangana neprovider
• Vertex Anthropic: Prompt caching ine cache control

Google Gemini: cachedContents API

Google Gemini inopa kutungamirira cache zvakajeka kuburikidza ne cachedContents API.

Magadzirirwo

$config = [
    'provider' => 'google-gemini',
    'model' => 'gemini-2.0-flash',
    'caching' => [
        'enabled' => true,
        'ttl' => 3600, // 1 hour in seconds
        'max_tokens' => 1000000, // Max tokens to cache
    ],
];

Kugadzira Prompt Yakachengetedzwa muCache

use Superdav\AI\Providers\GoogleGemini;

$gemini = new GoogleGemini( $config );

$cached_content = $gemini->create_cached_content(
    [
        'system_prompt' => 'You are a helpful assistant...',
        'context' => 'Large context document...',
        'ttl' => 3600,
    ]
);

// Returns: ['cache_id' => 'abc123', 'expires_at' => timestamp]

Kushandisa Prompt Yakachengetedzwa muCache

$response = $gemini->generate(
    [
        'cache_id' => 'abc123',
        'prompt' => 'User question here',
    ]
);

Cache Lifecycle

// List cached contents
$caches = $gemini->list_cached_contents();

// Get cache details
$cache = $gemini->get_cached_content( 'abc123' );

// Extend cache TTL
$gemini->update_cached_content(
    'abc123',
    ['ttl' => 7200] // Extend to 2 hours
);

// Delete cache
$gemini->delete_cached_content( 'abc123' );

Maitiro Akanakisisa eGemini

• Isa TTL yakakodzera: Enzanisa kuchengetedza mutengo ne cache staleness
• Chengetedza system prompts mu cache: Shandisazve system prompt imwe chete pazvikumbiro
• Tarisa mashandisirwo e cache: Tevera kuti ma cache api anoshandiswa zvakanyanya
• Bvisa ma cache akapera nguva: Nguva nenguva bvisa ma cache asingachashandiswi

Azure OpenAI: Prompt Caching

Azure OpenAI inotsigira prompt caching ine kutungamirirwa kweTTL otomatiki.

Magadzirirwo

$config = [
    'provider' => 'azure-openai',
    'model' => 'gpt-4-turbo',
    'api_version' => '2024-08-01-preview',
    'caching' => [
        'enabled' => true,
        'cache_control' => 'max_age=3600',
    ],
];

Kugonesa Caching

use Superdav\AI\Providers\AzureOpenAI;

$azure = new AzureOpenAI( $config );

$response = $azure->generate(
    [
        'system_prompt' => 'You are a helpful assistant...',
        'context' => 'Large context document...',
        'prompt' => 'User question here',
        'cache_control' => 'max_age=3600',
    ]
);

// Response includes cache usage:
// [
//     'content' => '...',
//     'cache_creation_input_tokens' => 1000,
//     'cache_read_input_tokens' => 500,
// ]

Cache Headers

Azure OpenAI inoshandisa HTTP headers pakutonga cache:

Cache-Control: max_age=3600

Mivalue inotsigirwa:

• max_age=<seconds>: Chengetedza mu cache kwenguva yakatarwa
• no_cache: Usachengetedze chikumbiro ichi mu cache
• no_store: Usachengetedze mu cache uye usashandisazve

Kuongorora Mashandisirwo eCache

$response = $azure->generate( [...] );

$cache_tokens = $response['cache_creation_input_tokens'] ?? 0;
$cache_hits = $response['cache_read_input_tokens'] ?? 0;

echo "Cache creation: $cache_tokens tokens\n";
echo "Cache hits: $cache_hits tokens\n";

Maitiro Akanakisisa eAzure OpenAI

• Shandisa maprompt anoenderana: Maprompt akafanana anobatsirwa ne caching
• Isa TTL ine mwero: Enzanisa mutengo nekutsva
• Tarisa cache metrics: Tevera kugadzirwa kwe cache zvichienzaniswa ne hits
• Unganidza zvikumbiro zvakafanana: Ronga zvikumbiro pamwe chete kuti uwedzere cache hits

OpenRouter: Caching Yakanangana neProvider

OpenRouter inotsigira caching kuburikidza neproviders vari pasi payo (OpenAI, Anthropic, nezvimwe).

Magadzirirwo

$config = [
    'provider' => 'openrouter',
    'model' => 'openai/gpt-4-turbo',
    'caching' => [
        'enabled' => true,
        'provider_cache' => 'openai', // Use OpenAI's caching
    ],
];

Kushandisa OpenRouter Caching

use Superdav\AI\Providers\OpenRouter;

$router = new OpenRouter( $config );

$response = $router->generate(
    [
        'system_prompt' => 'You are a helpful assistant...',
        'context' => 'Large context document...',
        'prompt' => 'User question here',
        'cache_control' => 'max_age=3600',
    ]
);

Sarudzo Dzakanangana neProvider

Providers vakasiyana vane nzira dzakasiyana dze caching:

// OpenAI-compatible caching
$response = $router->generate(
    [
        'model' => 'openai/gpt-4-turbo',
        'cache_control' => 'max_age=3600',
    ]
);

// Anthropic-compatible caching
$response = $router->generate(
    [
        'model' => 'anthropic/claude-3-opus',
        'cache_control' => [
            'type' => 'ephemeral',
            'max_tokens' => 1000000,
        ],
    ]
);

Maitiro Akanakisisa eOpenRouter

• Ziva caching yeprovider wako: Provider mumwe nemumwe ane nzira dzakasiyana
• Edza maitiro e caching: Simbisa kuti caching inoshanda neprovider wawasarudza
• Tarisa mitengo: Tevera kuchengetedza kunobva ku caching
• Shandisa models dzinoenderana: Kuchinja models kunoputsa cache hits

Vertex Anthropic: Prompt Caching ine Cache Control

Vertex Anthropic (Google Cloud) inotsigira prompt caching ine cache control yakajeka.

Magadzirirwo

$config = [
    'provider' => 'vertex-anthropic',
    'model' => 'claude-3-opus',
    'project_id' => 'your-gcp-project',
    'region' => 'us-central1',
    'caching' => [
        'enabled' => true,
        'cache_control' => [
            'type' => 'ephemeral',
            'max_tokens' => 1000000,
        ],
    ],
];

Kushandisa Vertex Anthropic Caching

use Superdav\AI\Providers\VertexAnthropic;

$vertex = new VertexAnthropic( $config );

$response = $vertex->generate(
    [
        'system_prompt' => 'You are a helpful assistant...',
        'context' => 'Large context document...',
        'prompt' => 'User question here',
        'cache_control' => [
            'type' => 'ephemeral',
            'max_tokens' => 1000000,
        ],
    ]
);

// Response includes cache metrics:
// [
//     'content' => '...',
//     'usage' => [
//         'input_tokens' => 1000,
//         'cache_creation_input_tokens' => 500,
//         'cache_read_input_tokens' => 300,
//     ],
// ]

Mhando dzeCache Control

• ephemeral: Cache kwenguva yechikumbiro (default)
• persistent: Cache muzvikumbiro zvakawanda (kana zvichitsigirwa)

Kuongorora Kushandiswa kweCache

$response = $vertex->generate( [...] );

$usage = $response['usage'];
$cache_created = $usage['cache_creation_input_tokens'] ?? 0;
$cache_read = $usage['cache_read_input_tokens'] ?? 0;

echo "Cache created: $cache_created tokens\n";
echo "Cache read: $cache_read tokens\n";

Maitiro Akanakisisa eVertex Anthropic

• Shandisa ephemeral caching: Zvakanaka pakucaching kwechikamu chimwe chete
• Seta max_tokens zvakakodzera: Enzanisa saizi yeCache nemutengo
• Ongorora cache metrics: Tevera kushanda kwecache
• Edza nebasa rako: Simbisa kuti caching inobatsira mashandisirwo ako

Zano reCaching Rinoyambuka Vapi

Kugadziriswa Kwakabatana

$config = [
    'caching' => [
        'enabled' => true,
        'default_ttl' => 3600,
        'providers' => [
            'google-gemini' => [
                'ttl' => 3600,
                'max_tokens' => 1000000,
            ],
            'azure-openai' => [
                'cache_control' => 'max_age=3600',
            ],
            'vertex-anthropic' => [
                'cache_control' => [
                    'type' => 'ephemeral',
                    'max_tokens' => 1000000,
                ],
            ],
        ],
    ],
];

Kuziva Mupi

$provider = $config['provider'];

$cache_config = $config['caching']['providers'][ $provider ]
    ?? $config['caching'];

// Use provider-specific caching configuration

Zano reFallback

try {
    // Try caching with primary provider
    $response = $primary_provider->generate( $request );
} catch ( CacheException $e ) {
    // Fall back to non-cached request
    $response = $primary_provider->generate(
        array_merge( $request, ['cache_control' => 'no_cache'] )
    );
}

Kuderedza Mutengo

Verenga Kuchengetedza Mari

$cache_created_tokens = $response['cache_creation_input_tokens'] ?? 0;
$cache_read_tokens = $response['cache_read_input_tokens'] ?? 0;
$regular_tokens = $response['input_tokens'] ?? 0;

// Typical pricing (varies by provider):
$cache_creation_cost = $cache_created_tokens * 0.00001; // 10x cheaper
$cache_read_cost = $cache_read_tokens * 0.000001; // 100x cheaper
$regular_cost = $regular_tokens * 0.00001;

$total_cost = $cache_creation_cost + $cache_read_cost + $regular_cost;
$savings = ($regular_tokens * 0.00001) - $total_cost;

echo "Estimated savings: \$$savings\n";

Mazano Ekugadzirisa

• Cache masystem prompts makuru: Kuchengetedza mari kukuru
• Shandisazve context: Cache magwaro econtext anoshandiswa kazhinji
• Batch zvikumbiro: Unganidza zvikumbiro zvakafanana kuti uwedzere cache hits
• Ongorora kushanda kwecache: Tevera mari chaiyo yachengetedzwa
• Gadzirisa TTL: Enzanisa mutengo nekutsva kwezviri muCache

Kugadzirisa Matambudziko

Cache haisi kushandiswa

• Simbisa kuti caching yakagoneswa mukugadziriswa
• Tarisa kuti maprompts akafanana (caching inoda kufanana chaiko)
• Simbisa kuti cache haisati yapera nguva
• Tarisa miganho yeCache inoenderana nemupi

Kugadzirwa kweCache kuri kutadza

• Simbisa kuti saizi yeCache iri mukati memiganho yemupi
• Tarisa kuti syntax yeCache control yakarurama
• Ita shuwa kuti mupi anotsigira caching yemodel yako
• Wongorora magwaro emupi kuti uone miganho

Mitengo isingatarisirwi

• Ongorora kugadzirwa kweCache uchienzanisa necache read tokens
• Simbisa kuti cache iri kunyatsoshandiswa
• Tarisa cache misses dzinokonzerwa nekusiyana kwemaprompts
• Funga kugadzirisa TTL kana zano reCache

Kuenzanisa Vapi

Chinhu	Gemini	Azure OpenAI	OpenRouter	Vertex Anthropic
Cache API	cachedContents	HTTP headers	Zvinoenderana nemupi	Cache control
Kudzora TTL	Zvakajeka	Kuburikidza nemaheaders	Zvinoenderana nemupi	Ephemeral/persistent
Saizi yepamusoro yeCache	1M tokens	Zvinoenderana nemupi	Zvinoenderana nemupi	1M tokens
Kuderedzwa kwemutengo	90%	90%	Zvinoenderana nemupi	90%
Kuongorora	Kwakadzama	Kuburikidza nemametrics	Zvinoenderana nemupi	Kuburikidza nekushandiswa

Matanho Anotevera

1. Sarudza mupi wako: Sarudza zvichienderana nezvaunoda
2. Gadzirisa caching: Gadzira caching inoenderana nemupi
3. Edza caching: Simbisa kuti inoshanda nemaprompts ako
4. Ongorora kushandiswa: Tevera cache hits uye mari yachengetedzwa
5. Gadzirisa: Gadzirisa TTL nezano reCache zvichienderana nezvabuda