Ukugcinwa kwee-Prompt kwi-Cache okuQaphela uMboneleli

Superdav AI Agent v1.12.0 yazisa ukugcinwa kwee-prompt kwi-cache okuqaphela umboneleli, okulungiselela iindleko ze-API kunye nokulibaziseka ngokugcina ii-prompt kwi-cache kubaboneleli be-LLM abohlukeneyo. Umboneleli ngamnye uneendlela kunye noqwalaselo lokugcina kwi-cache olwahlukileyo.

Isishwankathelo

Ukugcinwa kwee-prompt kwi-cache kukuvumela ukuba:

• Ugcine ii-prompt ezinkulu, ezisetyenziswa rhoqo kwi-cache
• Unciphise iindleko ze-API ngokuphepha ukuqhubekekisa okuphindaphindwayo
• Uphucule ukulibaziseka kwizicelo ezigcinwe kwi-cache
• Ulawule umjikelo wobomi be-cache ngokucacileyo

Ababoneleli abohlukeneyo baphumeza ukugcinwa kwi-cache ngeendlela ezohlukeneyo:

• Google Gemini: cachedContents API
• Azure OpenAI: Ukugcinwa kwee-prompt kwi-cache nge-TTL
• OpenRouter: Ukugcinwa kwi-cache okukhethekileyo kumboneleli
• Vertex Anthropic: Ukugcinwa kwee-prompt kwi-cache ngolawulo lwe-cache

Google Gemini: cachedContents API

Google Gemini ibonelela ngolawulo lwe-cache olucacileyo nge-cachedContents API.

Uqwalaselo

$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
    ],
];

Ukwenza i-Prompt eGcinwe kwi-Cache

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]

Ukusebenzisa i-Prompt eGcinwe kwi-Cache

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

Umjikelo woBomi be-Cache

// 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' );

Iindlela eziNgcono ze-Gemini

• Seta i-TTL efanelekileyo: Linganisa ukonga iindleko xa kuthelekiswa nokuphelelwa lixesha kwe-cache
• Gcina ii-prompt zesistim kwi-cache: Phinda usebenzise i-prompt yesistim efanayo kuzo zonke izicelo
• Beka esweni ukusetyenziswa kwe-cache: Landela ukuba zeziphi ii-cache ezisetyenziswa kakhulu
• Coca ii-cache eziphelelwe lixesha: Cima ngamaxesha athile ii-cache ezingasetyenziswanga

Azure OpenAI: Ukugcinwa kwee-Prompt kwi-Cache

Azure OpenAI ixhasa ukugcinwa kwee-prompt kwi-cache ngolawulo lwe-TTL oluzenzekelayo.

Uqwalaselo

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

Ukuvula Ukugcinwa kwi-Cache

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,
// ]

Izihloko ze-Cache

Azure OpenAI isebenzisa izihloko ze-HTTP kulawulo lwe-cache:

Cache-Control: max_age=3600

Amaxabiso axhaswayo:

• max_age=<seconds>: Gcina kwi-cache ixesha elichaziweyo
• no_cache: Ungasigcini esi sicelo kwi-cache
• no_store: Ungasigcini kwi-cache kwaye ungaphindi usisebenzise

Ukubeka esweni Ukusetyenziswa kwe-Cache

$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";

Iindlela eziNgcono ze-Azure OpenAI

• Sebenzisa ii-prompt ezingaguquguqukiyo: Ii-prompt ezifanayo ziyazuza ekugcinweni kwi-cache
• Seta i-TTL esengqiqweni: Linganisa iindleko xa zithelekiswa nokutsha
• Beka esweni iimetriki ze-cache: Landela ukwenziwa kwe-cache xa kuthelekiswa nokufunyanwa kwi-cache
• Qokelela izicelo ezifanayo: Dibanisa izicelo ukwandisa ukufunyanwa kwi-cache

OpenRouter: Ukugcinwa kwi-Cache okuKhethekileyo kuMboneleli

OpenRouter ixhasa ukugcinwa kwi-cache ngababoneleli abangaphantsi (OpenAI, Anthropic, njalo njalo).

Uqwalaselo

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

Ukusebenzisa Ukugcinwa kwi-Cache kwe-OpenRouter

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',
    ]
);

Iinketho eziKhethekileyo kuMboneleli

Ababoneleli abohlukeneyo baneendlela ezohlukeneyo zokugcina kwi-cache:

// 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,
        ],
    ]
);

Iindlela eziNgcono ze-OpenRouter

• Yazi ukugcinwa kwi-cache komboneleli wakho: Umboneleli ngamnye uneendlela ezohlukeneyo
• Vavanya ukuziphatha kokugcinwa kwi-cache: Qinisekisa ukuba ukugcinwa kwi-cache kusebenza nomboneleli omkhethileyo
• Beka esweni iindleko: Landela ukonga okuvela ekugcinweni kwi-cache
• Sebenzisa iimodeli ezingaguquguqukiyo: Ukutshintsha iimodeli kuqhawula ukufunyanwa kwi-cache

Vertex Anthropic: Ukugcinwa kwee-Prompt kwi-Cache ngoLawulo lwe-Cache

Vertex Anthropic (Google Cloud) ixhasa ukugcinwa kwee-prompt kwi-cache ngolawulo lwe-cache olucacileyo.

Uqwalaselo

$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,
        ],
    ],
];

Ukusebenzisa Ukugcinwa kwi-Cache kwe-Vertex Anthropic

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,
//     ],
// ]

Iintlobo zoLawulo lwe-Cache

• ephemeral: Gcina kwi-cache ngexesha lesicelo (okungagqibekanga)
• persistent: Gcina kwi-cache kuzo zonke izicelo ezininzi (ukuba kuyaxhaswa)

Ukubeka esweni Ukusetyenziswa kwe-Cache

$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";

Iindlela eziNgcono ze-Vertex Anthropic

• Sebenzisa ukugcinwa kwi-cache okwethutyana: Kulungile ukugcinwa kwi-cache kweseshoni enye
• Seta i-max_tokens ngokufanelekileyo: Linganisa ubungakanani be-cache xa kuthelekiswa neendleko
• Beka esweni iimetriki ze-cache: Landela ukusebenza kwe-cache
• Vavanya ngomthwalo wakho womsebenzi: Qinisekisa ukuba ukugcinwa kwi-cache kuyanceda imeko yakho yokusetyenziswa

Icebo lokuGcina kwi-Cache eliNqumla Ababoneleli

Uqwalaselo oluManyeneyo

$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,
                ],
            ],
        ],
    ],
];

Ukuchongwa koMboneleli

$provider = $config['provider'];

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

// Use provider-specific caching configuration

Icebo lokuBuyela Umva

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'] )
    );
}

Ukulungiselela Iindleko

Bala uKonga

$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";

Iingcebiso zokuLungiselela

• Gcina ii-prompt zesistim ezinkulu kwi-cache: Olona konga lweendleko lukhulu
• Phinda usebenzise umxholo: Gcina kwi-cache amaxwebhu omxholo asetyenziswa rhoqo
• Qokelela izicelo: Dibanisa izicelo ezifanayo ukwandisa ukufunyanwa kwi-cache
• Beka esweni ukusebenza kwe-cache: Landela ukonga kwangempela
• Lungisa i-TTL: Linganisa iindleko xa zithelekiswa nokutsha

Ukusombulula Iingxaki

I-cache ayisetyenziswa

• Qinisekisa ukuba ukugcinwa kwi-cache kuvuliwe kuqwalaselo
• Jonga ukuba ii-prompt ziyafana (ukugcinwa kwi-cache kufuna ukuhambelana okuchanekileyo)
• Qinisekisa ukuba i-cache ayikaphelelwa lixesha
• Jonga imida ye-cache ekhethekileyo kumboneleli

Ukwenziwa kwe-cache kuyasilela

• Qinisekisa ukuba ubungakanani be-cache bungaphakathi kwemida yomboneleli
• Jonga ukuba isintaksi solawulo lwe-cache sichanekile
• Qinisekisa ukuba umboneleli uxhasa ukugcinwa kwi-cache kwimodeli yakho
• Phonononga uxwebhu lomboneleli malunga nemida

Iindleko ezingalindelekanga

• Beka esweni ukwenziwa kwe-cache xa kuthelekiswa neetokeni zokufunda kwi-cache
• Qinisekisa ukuba i-cache iyasetyenziswa ngokwenene
• Jonga ukuphoswa kwe-cache ngenxa yokwahluka kwee-prompt
• Cinga ngokulungisa i-TTL okanye icebo le-cache

Uthelekiso lwaBaboneleli

Uphawu	Gemini	Azure OpenAI	OpenRouter	Vertex Anthropic
Cache API	cachedContents	Izihloko ze-HTTP	Okukhethekileyo kumboneleli	Ulawulo lwe-cache
Ulawulo lwe-TTL	Olucacileyo	Ngezihloko	Kuxhomekeke kumboneleli	Okwethutyana/okuqhubekayo
Obona bungakanani bukhulu be-cache	Iitokeni eziyi-1M	Kuxhomekeke kumboneleli	Kuxhomekeke kumboneleli	Iitokeni eziyi-1M
Ukuncitshiswa kweendleko	90%	90%	Kuxhomekeke kumboneleli	90%
Ukubeka esweni	Okuneenkcukacha	Ngeemetriki	Kuxhomekeke kumboneleli	Ngokusetyenziswa

Amanyathelo aLandelayo

1. Khetha umboneleli wakho: Khetha ngokusekelwe kwiimfuno zakho
2. Qwalasela ukugcinwa kwi-cache: Misela ukugcinwa kwi-cache okukhethekileyo kumboneleli
3. Vavanya ukugcinwa kwi-cache: Qinisekisa ukuba kusebenza nee-prompt zakho
4. Beka esweni ukusetyenziswa: Landela ukufunyanwa kwi-cache kunye nokonga iindleko
5. Lungiselela: Lungisa i-TTL kunye necebo le-cache ngokusekelwe kwiziphumo