Struct CacheConfig 

pub struct CacheConfig {
    pub enabled: bool,
    pub ttl: Option<CacheTtl>,
    pub max_breakpoints: Option<u8>,
}

Caller-facing control over provider-side prompt caching.

This is additive: a ChatRequest with cache = None preserves each provider’s default caching behaviour. Set it to shape (or disable) caching:

• enabled = false opts the request out of caching entirely — providers send no cache_control breakpoints.
• ttl selects the cache retention window (Anthropic ephemeral TTL).
• max_breakpoints caps how many cache breakpoints the provider may emit, in decreasing order of prefix stability (tools, then system, then the conversation tail). None leaves the provider’s default count.

Providers without a prompt-cache control ignore every field gracefully.

Fields

enabled: bool

Whether prompt caching is enabled for this request.

ttl: Option<CacheTtl>

Optional cache retention window. None uses the provider default.

max_breakpoints: Option<u8>

Optional cap on the number of cache breakpoints the provider emits.

Implementations

impl CacheConfig

pub const fn enabled() -> CacheConfig

An enabled cache config with provider defaults (no TTL override, all breakpoints).

pub const fn disabled() -> CacheConfig

A config that opts the request out of provider-side caching.

pub const fn with_ttl(self, ttl: CacheTtl) -> CacheConfig

Set the cache retention window.

pub const fn with_max_breakpoints(self, max_breakpoints: u8) -> CacheConfig

Cap the number of cache breakpoints the provider may emit.

Trait Implementations

impl Clone for CacheConfig

fn clone(&self) -> CacheConfig

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for CacheConfig

fn fmt(&self, f: &mut Formatter<'_>) -> Result<(), Error>

Formats the value using the given formatter.

impl Default for CacheConfig

fn default() -> CacheConfig

Returns the “default value” for a type.