## LLM cache


## llmcache.LlmCacheMode


How [CachingTransport](llm-cache.md#athome.llmcache.CachingTransport) treats each request against the store.


Usage

``` python
llmcache.LlmCacheMode()
```


## llmcache.LlmCacheSettings


The `[llmcache]` section: default cache mode and key schema version.


Usage

``` python
llmcache.LlmCacheSettings()
```


## llmcache.CachingTransport


Record/replay layer keyed on sha256(schema_version, method, scheme, host, port, path, sorted-query, body).


Usage

``` python
llmcache.CachingTransport(
    inner,
    mode,
    schema_version,
    db_path,
    lock=anyio.Lock(),
    stack=AsyncExitStack(),
    store=None
)
```


Headers are excluded from the key by design (auth and session noise never perturb replay), and only `2xx` responses are recorded. Streaming responses are fully buffered: the body is read into memory before it is stored. The sqlite store lives under `load(AthomeSettings).cache_root / "llmcache" / "llmcache.db"` and opens lazily on the first request, closing when the transport does.


#### Parameter Attributes


`inner: httpx.AsyncBaseTransport`  

`mode: `<a href="llm-cache.html#athome.llmcache.LlmCacheMode" class="gdls-link gdls-code"><code>LlmCacheMode</code></a>  

`schema_version: int`  

`db_path: Path`  

`lock: anyio.Lock = anyio.Lock()`    

`stack: AsyncExitStack = AsyncExitStack()`    

`store: `<a href="store.html#athome.store.Store" class="gdls-link gdls-code"><code>Store</code></a>` | None = None`  


## llmcache.RetryTransport


Retries 429/5xx/transport errors with capped exponential backoff (default 3 tries).


Usage

``` python
llmcache.RetryTransport(inner)
```


#### Parameter Attributes


`inner: httpx.AsyncBaseTransport`  


## llmcache.transport()


Build the record-replay transport stack for an `httpx.AsyncClient`.


Usage

``` python
llmcache.transport(*, mode=None)
```


#### Parameters


`mode: `<a href="llm-cache.html#athome.llmcache.LlmCacheMode" class="gdls-link gdls-code"><code>LlmCacheMode</code></a>` | None = None`  
The cache mode; `None` reads `LlmCacheSettings.mode`.


#### Returns


`httpx.AsyncBaseTransport`  
`CachingTransport(RetryTransport(httpx.AsyncHTTPTransport()))`, or a bare

[RetryTransport](llm-cache.md#athome.llmcache.RetryTransport) when the mode is `LlmCacheMode.BYPASS`.


#### Example

``` python
>>> AsyncOpenAI(http_client=httpx.AsyncClient(transport=transport()))
```


## llmcache.LlmCacheMiss


Raised in `LlmCacheMode.REPLAY` when a request has no recorded response.


Usage

``` python
llmcache.LlmCacheMiss()
```