Struct ContextTier 

#[non_exhaustive]
pub struct ContextTier {
    pub above_prompt_tokens: u64,
    pub input_per_million: f64,
    pub output_per_million: f64,
    pub cache_read_per_million: f64,
    pub cache_write_per_million: f64,
}

Higher rates charged above a context threshold.

OpenAI prices gpt-5.5 at $5/$30 below 272K prompt tokens and $10/$45 above it, and the GPT-6 models at 2x input and cache rates and 1.5x output above the same line — published as columns on the same pricing row, which is easy to miss if you go looking for a second row. A flat CostConfig under-bills those requests by 2x on input, and this crate’s whole compaction subsystem exists to run agents at high context, so the case is central rather than exotic.

The threshold is compared against the request’s prompt tokens — input + cache_read + cache_write — not the total including output.

Fields (Non-exhaustive)

Non-exhaustive structs could have additional fields added in future. Therefore, non-exhaustive structs cannot be constructed in external crates using the traditional Struct { .. } syntax; cannot be matched against without a wildcard ..; and struct update syntax will not work.

above_prompt_tokens: u64

Prompt tokens above which the tier rates apply.

input_per_million: f64

output_per_million: f64

cache_read_per_million: f64

Cache-hit rate above the threshold. 0.0 bills at this tier’s input_per_million, as for CostConfig::cache_read_per_million.

cache_write_per_million: f64

Cache-write rate above the threshold. 0.0 bills at this tier’s input_per_million, as for CostConfig::cache_write_per_million.

Implementations

impl ContextTier

pub fn new( above_prompt_tokens: u64, input_per_million: f64, output_per_million: f64, ) -> Self

A tier’s threshold and the two rates every tier has.

Cache rates are added with with_cache_read and with_cache_write, mirroring CostConfig::new and for the same reason. A cache rate left unset bills at this tier’s input rate (see CostConfig).

above_prompt_tokens is exclusive: a prompt exactly at the threshold stays on the band below, matching how vendors publish >272K.

pub fn with_cache_read(self, per_million: f64) -> Self

Rate for reading a cached prompt prefix above this threshold.

pub fn with_cache_write(self, per_million: f64) -> Self

Rate for writing a prompt prefix into the cache above this threshold.

pub fn is_configured(&self) -> bool

Whether this tier sets any rate. Mirrors CostConfig::is_configured.

Trait Implementations

impl Clone for ContextTier

fn clone(&self) -> Self

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ContextTier

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for ContextTier

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl PartialEq for ContextTier

fn eq(&self, other: &Self) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl Serialize for ContextTier

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl StructuralPartialEq for ContextTier